All that glisters

I've written often about scams, of one kind or another. I find them fascinating.

I've tied myself in knots classifying this one, though. To my mind most examples of it are clearly over the "scam" line, and I think almost everybody would agree that at least some examples of it definitely are, but...

Look. Here's the deal.

The other day, I wrote about "liquid metal bullion"...

Melted fusible metal in plastic bag

...which was presented on eBay as some kind of investment that's fun for all the family. It's actually of some interest as a novelty, but has little monetary value and is full of poisonous heavy metals.

While exploring the peculiar world of the "liquid bullion" dealers, I discovered another odd category of eBay "bullion":

"Gold" bars and coins, that actually have very close to no gold in them.

And "silver" ones, too, and a few others plated with more exotic precious metals. But mainly gold.

Small fake gold bar with a winged Jesus on it for some reason

I bought one. Here it is. I paid a grand total of $US2.30 for it, including delivery, from this dealer.

It took a couple of weeks longer to arrive than it should have - possibly because the sender didn't know the difference between Australia and the UK, as far as address labels go - but apart from that, the transaction was entirely unremarkable.

I was going to cut into the bar to show it wasn't solid gold, but since it sticks to a magnet, I think we can pretty much take that as read.

A metal that sticks to a magnet must contain iron, nickel and/or cobalt, iron being the cheapest. So under the plating this is clearly a slug of iron or steel of some sort.

For the sake of completeness, though, I still measured its vital statistics.

Fake gold bar

The bar's dimensions are about 44 by 28 by 3 millimetres, which would give it a volume of about 3.7 cubic centimetres if it didn't have rounded corners and that embossed image of Jesus-with-wings-for-some-reason on one side...

Fake gold bar in display case

...and an angel and some symbolic Commandments on the other. (It also came with this little clear plastic case, to help keep the practically molecularly thin gold layer intact.)

(Oh, and yes, I did specifically choose this particular style of object-of-no-value made to appear desirable by a perfunctory shiny coating. On account of the symbolism. I'm dead subtle, me.)

When I measured the volume of the bar more accurately via the immersion method (PDF), as per the liquid "bullion", I got 3.5 cubic centimetres.

When I weighed it normally, I got 27.4 grams.

That gives a density of 7.8 grams per cubic centimetre. My lab balance and cack-handed technique are accurate enough that I'd say with some confidence that the real density is somewhere in the 7.75 to 7.85 range.

The density of pure iron is 7.9 grams per CC; various steels have densities between 7.75 and 8.05 grams per CC. Common mild steel is about 7.85.

So yeah, this is indeed a chunk of cheap steel, as any fool who stuck a magnet to it could have told you without all the science stuff.

At this point you might be thinking, "No harm, no foul". It looks like gold, but it doesn't feel like gold or in any way beyond superficial appearance attempt to resemble gold. So it's just a decorative trinket, not an attempted scam. Right?

Well, maybe. Except the auction title was:

HOT EXTREMELY RARE!! "Jesus"_1 Troy oz. .999 24K Pure Gold Layered Bullion Bar

Let that sink in for a moment.

As I write this, the same seller has more bars just like this one, plus other ones with these descriptions:

Amazing price MAPLE LEAF GOLD BAR One Troy oz 100 MILLS .999 Gold 24K PLATED


1 oz 24K GOLD plated elephant OF SOUTH AFRICA the Krugerrand BAR 100 Mills RARE


...and so on.

All of the descriptions contain keywords you'll find in auctions of solid-gold items, but some of them also have the plain words "plated", and "gold clad ... iron". Others, though, only reveal their not-anything-like-solid-gold nature with odd terms like "100 Mills" or "gold layered".

Both of these terms seem to be recent inventions, at least when it comes to bullion. By definition, there's no such thing as "plated bullion"; it's as silly, though not as hazardous to health, as calling that low-melting-point alloy that has lead and cadmium in it "non-toxic".

EBay currently has quite a lot of allegedly-bullion items using these odd descriptions.

There actually is a unit called a "mil", with one L instead of two; it's a thousandth of an inch. That's obviously not what it means here, though, because a hundred mils is a tenth of an inch, which is 2.54 millimetres. If you can figure out a way to make something that's three millimetres thick yet plated with 2.54 millimetres of gold on both sides, a career as a TARDIS engineer awaits you.

What "mill" actually means to the eBay gold-plated bar-and-coin sellers is... unclear. Perhaps it's a millionth of an inch. A hundred millionths of an inch is 0.00254 millimetres, 2.54 microns; that actually does qualify per the US Federal Trade Commission as "Heavy Gold Electroplated". You can get thicker plating that that, too, up to the point where it qualifies as gold-filled, with the gold accounting for a readily measurable fraction of the item's weight, rather than just a barely-weighable plating. (Apparently a general rule of thumb for jewellery subject to wear is that one micron of plate thickness will wear off the item per year.)

Given that the "mill" is not any kind of defined unit and seems to be interchangeable with the similarly un-defined "gold layered", though, I don't think it's excessively uncharitable to assume that the actual thickness of the plating on these things is as thin as possible without letting the colour of the underlying metal show through.

I mean, the one I bought is supposed to be "1 Troy oz", too, but it only weighs 27.4 grams, not the 31.1 grams of an actual troy ounce. It doesn't even quite make it to an ordinary avoirdupois ounce; that's 28.35 grams. Given gross failures like this, I doubt the vendors spend a lot of time worrying about the actual thickness of their plating.

But so what, I hear you say. This is just the usual level of cheerful eBay flea-market dodginess, right? Anybody who's been on eBay for a while is probably familiar with its own special not-quite-scams.

Listings, for instance, that don't make it quite as clear as they might that the item being sold is an empty box which at one point contained the new and exciting game console prominently featured in the listing title. See also people selling a picture of a fancy guitar, or a miniature dollhouse version of a big-screen TV. Et cetera. If the buyer cannot figure out why a "one ounce" gold bar is selling for $2.30, wasting money on eBay is probably not their biggest problem.

I invite you, at this juncture, to check out the highest prices people have paid for "100 mills" or "gold layered" things on eBay, by searching completed listings. Red numbers indicate something that didn't sell, green numbers indicate a sale.

As I write this, that search is headed by "1 OZ GOLD SOUTH AFRICAN 2010 KRUGERRAND COIN BULLION 100 MILLS 999.9 24K LOT 10", which a UK seller unloaded for £670 ex delivery - more than a thousand US dollars.

Those were clearly not real Kruggerands, because the listing says: "This 2010 coin is layered with 100 mills thick of pure 24k Gold". But right before that, the listing copies from Wikipedia and says, "The Krugerrand is a South African Gold Coin, first minted in 1967 to help promote South African Gold. The coin, Produced by the South African Mint, proved popular and by 1980 the Krugerrand accounted for 90% of the global coin market".

Which is true. But those solid gold coins are not what this dealer is selling. They are selling ten coins that look a bit like them, but are each worth no more than my little plated Jesus-bar.

Unquestionably, the person who paid £677.95 delivered for these ten shiny poker chips was under the impression that they'd just bought ten ounces of highly fungible gold at a huge discount.

They are not alone in this thought. Scrolling down that search turns up a ten-gram "100 Mills" bar that sold for a hundred UK pounds, then a five-gram "100 Mills" bar selling here in Australia for $AU122.50, then a five-gram "100 Mills" bar from an Irish dealer selling for €87.50.

Four "2010 UK SOVEREIGN COINS -1oz - 24k PURE GOLD Layered .999 Fine -TAX FREE"? Those had "100 Mills" in the description, and went for £159.90 delivered. Another seller was pleased to relieve a customer of £154.94 delivered for "NEW 2013 Royal Coronation & 2012 Jubilee 24k PURE GOLD Layered Double Coin Set", again allegedly "100 mill" plated and "Genuine Coins - Not Copies Or Reproductions"!

That same seller also managed to unload a single "2010 BRITISH SOVEREIGN 24K PURE GOLD Layered Proof COIN -1oz .999 Fine *MUST SEE", for £106.99 delivered.

And on and on it goes.

So: Is this a scam?

I'd say yes, because "good faith" is a critical legal concept. Good-faith, as I've written before, is the undoing of a long list of "technically legal" rip-offs. If there is no way anybody would agree to a given deal if they knew exactly what it was, then camouflaging the true nature of that deal, however lightly, is attempted fraud.

Deals of this nature are, of course, not hard to find, and they're often being offered by large corporations, not eBay fly-by-nights. Payday-loan shops, dodgy mechanics, questionable sweepstakes, and umpteen outfits whose business model seems to accept a repeat-business level below one per cent and the kind of word-of-mouth goodwill usually only enjoyed by serial killers.

What about rebate programs that require you to send the same cut-out barcode from a package to two different addresses simultaneously? Reward-points programs predicated upon normal consumers' points expiring before they accumulate enough to be able to redeem them for anything? Airlines that routinely sell more tickets than there are seats on the plane, in the expectation that not everybody will actually get there on time (thanks, interminable "security" nonsense!)? And, of course, the worst invention in the history of capitalism, gift certificates, whose principal reason for existence is "breakage", that portion of the gift cards sold which are never redeemed.

There's plenty of other underhanded activity in the bullion market, too, with the endless promotion of overpriced "collectible" bullion coins (particularly to certain market segments...), and sharp dealing in the "cash for gold" business. But at least all of those outfits generally are selling and buying actual gold, not plated slugs that only superficially resemble actual bullion.

Advertising a near-worthless little chip of gilded steel as "Gold Bar 5 Grams 'Canadian Maple' 100 MILLS .999 24k Fine Bullion!" is not a good-faith act. You're clearly fishing for suckers.

There are some other murky terms used in describing these bullion-like shiny objects. "HGE", for "heavy gold electroplate", for instance, which is a term that exists in the jewelry market, but not so much in the bullion one. And "gold dipped", suggesting there's some worker out there spending all day dunking Krugerrand-resembling circles of steel in a cauldron of molten gold. There's "thick layer", too, which I think always indicates a layer actually notable for its thin-ness.

This rather cumbersome eBay search is for several of these terms, but not the slight-honesty-indicating terms "not pure gold", "not 100% solid gold" or "not solid gold". It has plenty of hits even when it's only searching the titles, and hundreds of hits if you click the little "Include description" box and then click Search again.

People sometimes pay big bucks even for the eBay items whose listings do include "not solid gold" disclaimers, though. And everybody who buys one of these things for more than the couple of bucks it's worth should have paid more attention. There's almost always some clue, if only what turns up when you search for terms like "gold layered" or "100 mills".

But not everybody is able to pay more attention, or aware of just how many scams there are on eBay.

I would also be willing to bet, given the long and depressing list of large green numbers in a completed-listings search for this gold-plated tat, that some people have spent a lot of money on these things. Perhaps they're hoping to quickly flip this amazing bargain to local precious-metal dealers. Perhaps they're under the impression that they're providing for their childrens' future. All they're actually doing, though, is transferring their life savings to a person selling scrap iron, and possibly lining themselves up for criminal charges if they ever try to sell these damn things on.

Not everybody selling gold-plated imitation bullion is a scumbag. Some of the "gold layered" listings are fixed-price "Buy It Now" items, for instance. Those cost a few bucks more than the auctioned ones usually sell for, but by their very existence they provide a strong clue that both they and their auctioned cousins aren't what your slightly dotty grandparent with an iPad and time on their hands might at first assume them to be.

Someone could still blow their entire retirement nest egg on the Buy-It-Now ones, but it'd take some effort. And the buyer would at least end up with a really big pile of almost worthless gold-plated novelties, which'd look good in an outraged local news story.

Some of these things also have pretty-much-honest descriptions, that clearly say something like "plated" or "replica" instead of "dipped" or "layered" or whatever. (The one I bought may not have put any disclaimers in the title, but its description text did contain "*PLEASE REMEMBER THESE BARS ARE NOT SOLID GOLD*".)

Even these better dealers do still love the magic word "bullion", but they're nonetheless more or less in "good-faith" territory, if you ask me. Even a moron in a hurry might realise the product is not solid gold when it says "plated" right in the auction title.

Oh, and just to confuse things even more, you can get real silver coins and bars that've been "layered" with gold. People overpay for those, too. As I write this, a Completed Listings search shows that someone thought a "2000 Washington Mint Sacagawea 24kt Gold Layered .99 Silver 4 Troy Oz Coin" was worth $US167.49 delivered. The gold value of that coin is as usual negligible, but presuming the seller's telling the truth about the amount of silver in it, then it is at current spot prices worth about eighty bucks.

So, still a rip-off, but only by about a factor of two.

("Silver-gilt" items are quite common in the legitimate jewelery business, especially for large items like sculptures and medals. Olympic "gold" medals, for instance, are silver-gilt to keep the price down. By specification they have to be be at least 60 by three millimetres, which at 2014 gold prices would make them cost the thick end of seven thousand dollars. There are more than three hundred events in a Summer Olympics and another hundred in the Winter, so that'd add up, especially for larger-than-spec medals; the London 2012 medals were unusually large, at 85 by 7mm. That much solid gold is currently worth well north of $30,000. So instead, Olympic golds are silver with a generous six-or-more grams of gold plated onto it, to make sure that even if the medal-winner insists on wearing the thing around all day, it won't wear through the plating.)

As I write this, a Completed Listings search for these gold-plated silver coins shows only six sold going back to May this year. The least anybody paid for one was $US105 delivered. For a coin worth, I remind you, $80. And that only when you manage to find someone who'll listen to your story about how there really is some silver there under the silly gold plating.

(This problem may solve itself, because very thin gold plated straight on top of silver will slowly turn silver and tarnish as silver atoms migrate through the gold. To avoid this, "proper" gold-plated silver jewellery has "barrier layers" in between, in a sandwich that may be silver, then copper, then nickel, then finally the gold. I doubt the sellers of "gold layered" "100 mills" silver coins go to these lengths to make sure their products retain their lustre.)

High in the most-expensive-first Completed Listings searches you'll also find a number of people paying a few hundred dollars for one hundred plated coins or ingots. Those people have not been ripped off either, though I presume most of them are hoping to get in on this occasionally-lucrative business themselves.

Please don't do that.

If you appreciate kitsch, do feel free to decorate your wall with the complete series of ULTRA RARE SOVIET NAZI JESUS ELEPHANT LUCKY MONKEY MARTIAN GOLD LAYERED ALMOST AN OUNCE HYPERBULLION INGOTS.

But I wouldn't pay more than two bucks a unit, if I were you.

A metallurgical detective story

There are people on eBay selling "liquid metal bullion".

They don't like to tell you exactly what metal this alleged "bullion" is, except they always swear it doesn't have any mercury in it.

They generally say it's solid at room temperature, but will melt from the heat of your hands.

Those of us with a mild recreational interest in the periodic table will draw a rapid conclusion if given the characteristics "metal, melts at body-heat temperatures, doesn't explode on contact with moisture...

...and non-toxic."

That'd have to be gallium, right?

Wrong. I bought some, and it's nothing like the gallium I already own.

I didn't know what the hell the "liquid bullion" was. Not, at least, until I played around with it for a while.

Fusible metal ingot

There seems to be some sort of tradition in the hobbyist low-melting-point-alloy business of casting your little ingots in unorthodox moulds. The mould is usually something that clearly indicates that the metal was liquid at temperatures low enough that to not instantly destroy a chocolate-box tray, silicone ice-cube tray, or similarly non-refractory mould material.

Wood's metal casting

I cast my own Wood's metal in a Lego mould.

You could craftily fake this by casting wax in a chocolate tray, then using that form to make a sand mould, or something, but I don't know of any such scandals in the retail weird-metal market.

All of the low-melting-point alloys exist because of the odd fact that mixtures of chemicals can have a lower melting point than any of the ingredients.

On the face of it, this doesn't make sense. I mean, the universe should be nice and sensible and line up with the way ancient philosophers hoped it worked, from tiny billiard-ball atoms all the way up to clockwork galaxies. Then, the melting point of an alloy would be the melting point of its constituents, weighted by what proportion of the alloy each constituent took up. So for Wood's metal, for instance, you'd have:

50% bismuth, melting point 271.5°C
26.7% lead, melting point 327.46°C
13.3% tin, melting point 231.93°C
10% cadmium, melting point 321.07°C

Weighting each of those by the fraction they take up gives 135.8, 87.4, 30.8 and 32.1; add those up to get your naïve simple mathematical logical melting point and you get 286.1°C.

The melting point of Wood's metal is actually only about 70°C. Stuff like this is why metallurgy was much more art than science for a long, long time.

(In case you're wondering, which you probably aren't but I was, for these kinds of calculations it's safe to use Fahrenheit, Celsius or Kelvin temperature scales. The arbitrary zero points of Fahrenheit and Celsius don't screw it up. Beware anybody who tries to tell you that a 30°C day is "twice as hot" as a 15°C day, though, because that's so dumb as to possibly be not even wrong. 15°C is 59°F, for instance; making 59°F "twice as hot" gives you 118°F, which is 47.8°C. Kelvin starts at absolute zero, so it's the only scale in which you could actually fairly say one temperature is twice another, though I'm still not sure how useful such an observation could be. Starting at -273.15°C makes "doubling" room temperature in Kelvin rather dramatic, though; 15°C is 288.15 Kelvin, double that is 576.3 K, which is 302.85°C.)

Many alloys don't have this oddly low melting point. Brass, for instance, has a melting point from about 900 to about 940°C depending on its formulation; it's composed of copper (melting point 1085°C) and zinc (melting point 420°C). The melting point of brass is higher than you'd expect from naïve proportion calculations.

But the most common low-melting-point alloy is ordinary tin-lead solder, which exhibits the melting-point-reduction effect. Tin melts at 232°C, lead melts at 327°C, but if you mix 63% tin with 37% lead you get an alloy that melts at only 183°C.

And so, back to my "liquid bullion" ingot, which I bought on eBay Australia for $AU19.01 delivered after watching several people buy their own for prices that exceeded my modest snipe.

It was quite small. Only about four centimetres in length...

Fusible metal ingot in hand

...and it weighed more than fifty grams.

That made it dense enough that, despite the seller's claims of non-toxicity, I treated it as if it were made of solid cadmium until I could figure out the thing's composition for myself.

(The seller was this guy - possibly NSFW! - who is now out of the "liquid bullion" business, having found the whole thing to be "nothing but a headache". That "NSFW" is there because after he got out of the liquid bullion business, he sold several pornographic coins. I am not making this up. As I write this he's only selling a sofa, but I'm sure he'll offer the Internet flea market some more eyebrow-raising products in the near future.)

The listing for my "bullion" ingot gave no hints regarding its makeup, but I bought it anyway, partly because low-melting-point metals interest me. I also figured that "liquid metal bullion" might be just as entertaining as "copper bullion", with which I had a lot of fun a few years ago.

(Tl;dr: Base metals sold by the troy ounce may be a fun novelty, but are not a good investment.)

Copper-bullion sellers are still rampant on eBay, but this liquid-metal schtick is new, and extraordinary.

It is, you see, mystery bullion! An unknown metal! Usually billed as very rare and valuable and desirable, whatever it is, but available to you today for amazing prices!!1!

I saved the listing from which I bought my little ingot. I won't upload the whole page-copy here, though, because malware-detection services tend to flip out, with some justification, if they find what looks like an eBay listing on some site other than eBay.

Here's what the listing said, though, with only the eBay trimmings and images removed.

Children, avert your eyes! Just reading this may delete more than a year of science education from your brain, and make you noticeably less intelligent than you were before!

999 Pure Liquid Metal Bullion INGOT 99c A GRAM 52.6 Gram Not 1oz Silver bar Coin



You will receive 1 ONE
Ingot as pictured which weighs 52.6 GRAMS IN TOTAL. Size is (1 & 3/4" long) (1" Wide)  (1/4" High)




Purchase as many as
you like. You will only pay the quoted postage cost as I will cover any EXTRA

This metal is
considered  bullion as it is pure rare

It melts at the low
temperature of less than 30 degrees, will even melt in your hand if held long


Spot Price is
increasing at a rapid rate with the demand of this

Pure Liquid Metal

Only 99 cents per
gram and one off low postage cost of $3.00

There's... I mean...

Look, I'm not even going to start with that description. It's the sort of scientific word-salad more often seen in explanations of crackpot cancer cures.

Oh, OK, just one thing: "Rare earths", and metals which are precious for their rarity, are not the same thing, no matter what certain eBay sellers think.

The elements known as "rare earths are actually quite common; the only "rare" thing about them is their concentration in any given load of ore, meaning you need to dig up a lot of the planet to get a little bit of rare-earth element. And then it's difficult to separate the different rare earths from each other, because several of them have very similar chemical properties.

Rare-earth elements are today used to make neodymium-iron-boron magnets, hence the term "rare-earth magnet". Modern lighter "flints" are made from a very sparky, pyrophoric...

...and somewhat excitable...

...alloy of different rare earths, plus a few other things and iron for strength.

Rare-earth magnets and lighter flints are not very expensive per gram, though, because they contain no precious metals. For a few bucks you can now buy a ferrocerium stick intended for use as an emergency fire-lighter - just scrape it with a knife blade, file or similar item to create a shower of sparks.

(There are fancy versions of these things with built-in scrapers, but a simple bare ferrocerium rod is almost as good. You can get a little one with a handle, perhaps a stick of magnesium too for use as high-temperature tinder, and a bit of hacksaw blade for scraping and spark-striking, for about a dollar delivered. A chunkier bare ferrocerium rod will only set you back a few bucks from a dealer who doesn't quite know the difference between magnesium and ferrocerium, and may theoretically save your life one day. It will definitely provide you with considerable entertainment and some tiny holes burned in whatever happens to be near you when you play with it.)

The YouTube link in the above exercise in eBay creative writing goes to this video, from the brain-polluting "HouseholdHacker". That dude used to make ridiculous practical-joke "how-to" videos, which on the one hand encouraged a lot of adults to do entertainingly silly things, but on the other hand probably turned some kids off science. Which took that guy right the hell off my Christmas-card list.

Now, though, HouseholdHacker seems to be producing serious videos. The one linked to by the liquid-bullion guy isn't what you'd call packed with educational information, but the only actual inaccuracy I noted in it was incorrect rounding so the melting point of gallium was 0.1°F off. That is not exactly a capital crime.

But I still think that you're going to transition from "joke videos to get people to do stupid things" to "actually telling the truth", you shouldn't keep your old name. Mixing the two is completely uncool, man.

(Oh, and while I'm on this subject, see also my favourite example of this latter crime. Good ol' Kip deleted all of his highly remunerative Metacafe videos at some point after he reinvented himself as the video face of Make magazine, thereby ensuring that I stopped watching any of their videos. I think Make came to their senses and quietly fired him after a year or three; their videos are much better now, and they've recently started an interesting new series.)

If you want a video about low-melting-point alloys that's not from a professional bullshit artist, you could do a lot worse than turn to "Brainiac75":

Oh, and if you want a good video about gallium alone, then you obviously need to turn to actual scientists...

...and their magnificent example of an actual scientist who looks like a mad one from a horror movie.

Aaaaanyway, anybody who hasn't yet died of old age reading this page may remember that the question was... what is this "liquid bullion" stuff?

While I was sniping auctions, little fifty-gram ingots like mine kept selling for twenty-five Australian dollars or more. That's a good price for fifty grams of gallium, but it's not a good one for a similar amount of toxic low-melting-point alloy. Small amounts of anything cost more per gram, but you don't have to buy a huge amount to pay a lot less. Brainiac75 above said he paid only ten Euro cents per gram for some of the lower-melting-point alloys in his videos.

(The very lowest-melting-point alloys in Brainiac75's video are alarming concoctions like an amalgam of periodic-table neighbours mercury and thallium. That is not ten cents per gram, but it stays liquid down to -60°C. Cesium-potassium-sodium makes it down to -78°C without solidifying, but it also explodes on contact with water.)

For my first attempt at identifying the metal, I contacted the seller thusly, batting my eyelids innocently:

I've received my little ingot, and now I find myself wondering what it's actually made of. Your listing doesn't mention this, other than to say that it contains no mercury. What actually IS this alloy?

I'd also be interested to learn where to look up the "spot price" you mention in the listing. (Which again, of course, requires me to know what alloy this actually is.)


While I waited for him to reply, I measured the little ingot's density.

Accurately calculating the density of a small object is tricky. Getting a vague ballpark figure isn't hard, especially if the object is roughly a rectangular prism, as this one was. Just measure the edges, fudge any bevelled edges into a sensible-looking in-between number, and then multiply the numbers. Doing that with the little ingot gave me a volume of about 5.6 cubic centimetres. Since it was bang on its advertised mass of 52.6 grams, this gave me a density of about 9.4 grams per cubic centimetres.

The density of solid gallium is only 5.91 grams per cc, so clearly this wasn't gallium.

(Gallium is also one of those odd materials that expands when it freezes; liquid gallium's density at its melting point of 29.8°C is about 6.1 grams per cc.)

My faithful triple-beam laboratory balance gives me quite accurate weight numbers, but I wanted a more accurate volume than fudged dimension-multiplication could offer.

When a metal has a low melting point you can, of course, just melt it and pour it into a graduated cylinder to measure its volume. But gallium, if there was any of it in this alloy, tends to "wet" a wide variety of other substances. So, presuming there was gallium or something that behaves like it in this alloy, getting all of it out of a narrow graduated cylinder again could be difficult.

Another way to measure volume is by filling a graduated container with water or oil or whatever else is compatible with the object whose volume you want to measure, and then dropping the object into it and seeing how far the water level rises. This often doesn't work any better than just measuring the edges, though. It's a good quick strategy for extremely irregular objects - figuring out this technique is what is suppose to have sent Archimedes running naked down the street shouting "Eureka!" - but I've tried it several times with different items, and every time I got miserably inaccurate results.

There's a much better way of measuring object volume by immersion, though. You just need to add a precision scale to your apparatus. Pretty-well-calibrated 0.1-gram-resolution digital scales are now commodity items, and my abovementioned lab balance will do the job nicely.

What you do is, you put some water - or, again, a different liquid if water is incompatible with the object you're measuring - in a vessel deep and wide enough to completely submerge the item whose volume you're measuring, without the object having to touch the bottom.

You then weigh the vessel and the water, or just press the zero-out "tare" button on your digital scale.

Now, you immerse the item you want to measure in the water. If it's less dense than water you have to push it down into the water until it's fully submerged, but it's probably more dense than water, in which case you can just suspend it rather than push it in.

The important part is that the object must be immersed, but not resting on the bottom of the container. This is because what you're measuring is the increase in weight of the container, not the rise in level of the liquid in it.

Whatever you suspend your object with should have as close to zero volume as you can manage. I used some kapton tape, partly because it is narrow and extremely thin yet has good adhesive, but mainly because it is unquestionably the scienciest of all of the more than two dozen kinds of tape I have to hand.

("Florists' crepe-paper tape?" Got that. "Colourful metallic tape less than a millimetre wide meant for decorating fingernails?" Yup. "Copper and aluminium foil tape?" Of course. "Self-amalgamating?" Which kind would you like, the old rubbery type or the new silicone stuff? "Gaff?" Multiple colours. "Foam draught-excluding door-seal tape?" Please. "Bendable-fridge-magnet tape?" Yes sir. PTFE thread-sealing tape? Naturally. "Unstretchable fibre-reinforced tape?" Ashamed to say I have only glass-reinforced, must get the aramid kind too. "Velcro tape and liquid tape?" Possibly the first and definitely the second doesn't really qualify as tape, but I've got 'em both anyway. And you can buy off-brand probably-kapton polyimide tape all over the place these days; it's generally just called "high temperature tape".)

Again, if you're measuring the volume of a ping-pong ball or something by the immersion method then you'll have to push it down into the water, but that'll still work. You could push it in with three needles mounted on some gantry over the scale, for instance.

Anyway, you suspend or shove the thing you're measuring into the water, suspending or shoving as little other stuff in there as possible, and the vessel will then become heavier by the mass of the liquid the object has displaced. Water weighs one gram per cubic centimetre at one gravity, so presuming you're using water and don't need numerous decimal points of accuracy, each gram of weight gained equals one cubic centimetre of object volume.

If you're now having some kind of "common sense" brain-spasm, wondering why a ping-pong ball shoved into a glass of water should make that glass as much heavier as would an identically-sized sphere of tungsten suspended in it, you may find this PDF soothing.

The initial mass of my glass plus water was 436 grams even; dangling the "bullion" ingot in it raised that to 441.5 grams, for a volume of 5.5 cubic centimetres.

This made me pleased about my original guesstimate of 5.6 cubic centimetres, though slightly less pleased about the time I'd spent bent over a laboratory scale to get a scarcely-different number. It's a bit like that story about how the Great Trigonometric Survey painstakingly measured the height of Mount Everest and came up with exactly 29,000 feet. That's exactly how tall everybody had always said the mountain was anyway, so, the story goes, they added another two feet to prevent people thinking they'd actually just gone to ground in a club in Calcutta and spent their time inventing snooker and the gin and tonic.

Anyway, 5.5 cubic centimetres and 52.6 grams gave me a density of 9.56 grams per cubic centimetre.

I now had a reply from the seller regarding what he reckoned I'd actually bought. He said:

Hi, the metal is frenchs metal type3 or gallium, both the same.


He was receptive to my then pointing out that "French's metal" and gallium are very much not the same, the latter being non-toxic and the former containing both lead and cadmium. It was at this point that he told me he wasn't selling this stuff any more on account of its headacheyness, which is I suppose one way of describing what happens when you sell poisonous heavy metals, both lead and rather more scary cadmium, as "non toxic and non hazardous to handle".

"French's metal" is an unusual term for an unpopular substance. It's easy to find people selling Wood's metal, which is bismuth, lead, tin and cadmium, and melts around 70°C. Rose's metal is also pretty commonplace; it's just bismuth, lead and tin, so not as poisonous as Wood's metal, and melts just below the boiling point of water.

French's metal winds the melting point down to only about 41.5°C by adding indium to the Wood's-metal mix. There are some further variants that melt even lower thanks to the presence of thallium as well; if this stuff really melted in your hand, I strongly suspect it'd have to be one of the thallium alloys.

Which would be bad. Especially if you were melting it in your hand.

There are very good reasons to have as little thallium in your life as possible. Cadmium is something in the order of ten times as toxic as lead, but you can at least touch the stuff with your bare hands without appreciable danger, provided you wash your hands thoroughly afterwards.

Metallic thallium can pass through the skin, though, and is much more toxic than cadmium. Exact comparisons are difficult, because human thallium exposure is usually via one of its several useful-yet-toxic compounds, rather than the pure metal. But thallium is probably tens, if not hundreds, of times as toxic as cadmium. See this PDF from the US EPA, for instance, and compare with MSDSes (previously) for cadmium, like this one or this PDF one.

You really, really don't want to get any thallium on you.

(One of the symptoms of thallium poisoning is that your hair falls out. Needless to say, this means thallium sulfate used to be used as a depilatory, not that long ago. See also the use of lead and mercury compounds for skin whitening. Thallium is also still used in some countries to poison rats, ants and troublesome spouses.)

Fortunately, most people don't need a fusible alloy that melts at as low a temperature as bismuth-lead-tin-cadmium-indium, and fewer people still need the alloys with thallium as well. Presumably, because of this relative unpopularity, "French's metal" and its relatives are often not called that, and just stuck on page 137 of the specialist-alloys catalogue with no name beside their ingredients and melting point.

On with the investigation, then. What actually is the melting point of this stuff?

If it were pure gallium then it would indeed melt in your hand, provided the ambient temperature was high enough; gallium melts at 30°C (86°F). It's too dense to be pure gallium, though, so if it melts at blood temperature then it's probably terrifyingly toxic.

"French's metal" formulations - without thallium - are frequently quoted as melting at 117°F, which is 47.2°C, way higher than any survivable body temperature. Similar alloys with added thallium are quoted as low as 105°F, which is 40.6°C and still not "body temperature" unless you're quite gravely ill. Measuring the melting point can therefore help me decide whether it's moderately-nasty French's metal or some handle-with-gloves thallium alloy.

I've got a high-accuracy digital probe thermometer, from back in the day when I reviewed incredible quantities of CPU coolers. (It's tempting to simplify the setup by just pointing one of today's inexpensive non-contact infra-red thermometers at whatever you're heating, but in this case that wouldn't work.)

So I set up the sort of advanced experimental apparatus for which I am so justly renowned...

Gimcrack apparatus for measuring melting point of fusible metal alloy

...with the metal ingot again suspended in water, but this time inside a resealable storage bag, of the type generically referred to as, but in this case not actually a, Ziploc.

The bag insulated the metal from the water, of course, and my temperature probe was in the water, not inside the bag to get all probably-cadmium-ed up. So I needed to be a bit crafty to get a useful melting-point number.

What we're interested here is how low the temperature the metal melts at is, not how high it is, if you get my meaning. So I ran the water temperature up to 50°C (122°F), at first. Then I turned the heat off and snapped the above picture of the setup, while the metal got around to melting.

Melted fusible metal in plastic bag

After taking this photo, I hung the melted metal in its bag back in the water, and allowed the water to cool.

As the water temperature fell through the low forties Celsius, the metal started solidifying again. Crystals started forming in the liquid, so at first the metal in the bag felt like a dense liquid with a little sand in it, then more and more like unusually heavy wet sand, until finally it solidified entirely.

I think this might mean this alloy is non-eutectic, with no clear melting point because different components melt at different temperatures. It could also just be the normal way a cooling metal will crystallise if you keep poking at it and examining its texture, though - the "liquidus temperature" is defined as the temperature at which solid crystals can coexist with melted material. I'm not sure.

The metal was wholly solid again when the water temperature was 40°C (104°F). Taking the bag's insulation effect into account, that told me the melting point was above 40°C and below 50°C, so the "melts in your hand" claim was clearly disproven, but I didn't yet have much idea exactly which alloy I was looking at.

I then ran the temperature slowly back up again, and the metal was re-melting, with the same sandy-liquid feel, by the time the water was back up to 47°C. But, notably, not when the water was only at 42 or 43°C, which would indicate a scary thallium alloy.

And then a pinhole opened in the corner of the bag and tiny droplets started escaping, and I terminated the experiment before I got heavy metals all over the kitchen again.

(Perhaps a genuine Ziploc® Brand bag would have been tougher. Squishing a gritty liquid with 83% the density of lead around in a the pointy corner of a polyethylene plastic intended to contain only food would probably cause any such bag to spring a leak, though.)

I could have re-bagged the metal and kept refining my temperature range, but what I'd done so far makes me confident that the melting point is somewhere in the 42-to-47-degree-C range, and probably the upper portion of that range. So I'm about 95% sure that this is indeed some kind of French's metal alloy containing lead and cadmium, but not deadly thallium.


If you want a relatively inexpensive fusible alloy to play with, go for Field's metal. It melts at about 62°C (144°F), and it contains only bismuth, indium and tin, so genuinely is non-toxic. Bismuth and indium are a bit expensive, which means Field's metal is too, but you could cast a teething ring out of it and probably not harm the baby.

Describing any of these low-melting-point fusible alloys as "bullion", though, is if anything even sillier than doing the same for copper. They're not worth enough per kilo to be an investment item, and most of them contain lead, cadmium and/or even thallium, which makes them less "valuable heirloom" and more "toxic waste".

Two low-melting-point metals, one assistive cat

Here, along with one helpful wubble, is where my own "liquid bullion" ended up. I've left it in the triangular shape the corner of the plastic bag gave it, along with the little spherical droplets that escaped into the saucepan. It's quite pretty, covered with tiny sparkling crystal surfaces; cooling it slower might have made bigger crystals, though nothing that could compete with bismuth.

To the right of the "bullion" lump is my sample of gallium, which is currently solid. Gallium is one of those substances that'll stay liquid below its freezing point if nothing serves as a nucleation point to start it crystallising. (The same thing can happen with water and various beverages in smooth plastic or glass containers).

Gallium sticks to almost everything, though, so if you slosh it around in the bottom of a container it'll make a silvery mirror out of whatever parts of the container-sides it touches. Once it finally decides to solidify - which, for my gallium at least, can take weeks - you can flick the flexible sides of the container to break the thin gallium coating off them. The result is what you see in the above picture - uneven coverage of the sides with thin plating I didn't manage to dislodge, and random dislodged flakes of gallium sitting on top of the solid layer in the bottom of the container.

(I rather like these little PET bottles, by the way. My gallium came from the Amazon seller in a tough grey translucent container that doesn't show it off nearly as well as this new one. Five of these eighty-millilitre containers, about 8cm high and 4.5cm wide, only cost me $AU3.88 delivered on eBay. They seem to be a couple of bucks more expensive now. UPDATE: But because they're standard PET bottles blown to shape from a preform, they shrink if you put them in boiling water! I think I can re-liquefy my gallium in one of these bottles, but now I've got one funny-looking one from pouring too-hot water on it.)

As I write this, the spot price of silver is less than $US20 per troy ounce; precious metals in general have taken a dive in the last few months. The spot price for gallium is at the moment maybe $US500 per kilogram, and one kilogram is 32.15 troy ounces. So gallium is something like $US15.55 per troy ounce, right up there with silver.

There is, just as with copper, no real liquid market (pun not intended) for small quantities of high-purity gallium. But the value of the stuff is sufficient that if you manage to buy it by the kilo at close to the bulk spot price, it really could qualify as an investment.

If you buy fifty grams of gallium in a little bottle from that Amazon dealer then you'll be paying a large markup on the bulk price, as is normal for metals other than the generally-accepted "precious" ones sold at retail in ounce quantities. It's also possible to quickly turn gold, silver or platinum into cash, if you suddenly need to. In a similar situation with gallium you'd have a hard time finding people who even know what it is, much less people who'll buy it from you at a fair price, in a hurry.

On the other hand, gallium's value is closely pegged to its real usefulness in the world. Gold, silver, platinum and palladium all have real-world uses, but their value is far higher than those uses justify. A large slice of the precious-metals market is people buying the stuff as an investment or just a store of value, perhaps as an alternative to a savings account in their shaky local currency. (India, in particular, has a strong tradition of storing household money in gold.)

Precious metals have never been a good long-term investment in the modern world, but they're portable and fungible, and that counts for a lot, even if you accept that you could make more money with index funds, bonds, or often even crappy-yield savings accounts.

Nobody's casting gallium ingots and keeping them in Fort Knox, though. Which is just as well, because the stuff would totally pull a Cryptonomicon if you turned the heating up too far.

A bottle of sloshy liquid non-toxic gallium is a lot more fun than a similar amount of similarly-valuable but much-easier-to-sell silver, though. I think that's a fair trade.

But don't buy weird "bullion" of any kind from eBay dealers, especially ones that say their product is non-toxic but aren't actually sure what it is. And if you are an eBay dealer selling weird "bullion", for pity's sake figure out what it is that you're selling, lest you be the next schmuck to put a "safe for kiddies!" sticker on a lump of cadmium. Or worse.

MAKING money for nothing and FINDING chicks for free

In the comments to my post about people who think that all I do is beg for money on the Internet, Fallingwater asked:

Do you think a single, low-expense person can actually, really make a living with a site such as yours in 2012? I'm talking referrals, sponsorships and such, not living off donations (not that I'd mind, but I think you'd need a Wikipedia-like amount of readers, and possibly Jimmy Wales' creepy face, to pull that off).

I think you definitely can, even without staring into the soul of everybody who visits one of the most popular Web sites ever.

(Should I decide to try that, I would of course use...

Daniel B. Nosemonster

...this picture.)

That doesn't mean it's easy to make money with a Web site these days, though.

The main problem is that there's no way for a review site or similar enterprise to make a decent amount of money from the beginning. If it's a review site, every review can make you a small but non-trivial amount of money for the first week or so of its life, and then long-tail off into cents per day. But if you've got a thousand pages each making you 15 cents a day, you'll be doing OK. When you've only made it to the 50-page mark, though, you could easily be grossing no more than 25 bucks a day, which ain't gonna pay the rent in most of the Western world.

If you're in Africa or eastern Europe or something then this could of course still be a very workable proposition, but making affiliate deals with local businesses, generally on a per-sale basis, is a major way for small sites to get going, and local businesses in Uganda have a lot less money to throw around. There may also be major obstacles to getting money from richer countries sent to you in a poor one; I don't know.

I have always had it very easy. This is partly because I was smarter with my money during the dot-com nonsense than some of my friends. (Shiny new car and inner-city apartment? Nope, I'll go with rusty used car and living with mum, thanks. I did blow a surprising amount of money on this toy, though - brushless motors were EXOTIC back then.)

My easy ride was also partly because I for some reason am good at writing, and at understanding computers.

(I think Michael Bywater was partly responsible for this. He wrote the computer column in Punch in the eighties, giving me the chance to read comedic writing about Lotus 1-2-3 when I was a small child with absolutely no understanding of what this software actually did, but he also anonymously wrote the gonzo-ish "Bargepole" column, which I also didn't really understand but which connected some of my neurons in quite novel ways.)

I've had it so easy mainly because I was lucky, in the abstract sense of being born white and male in a rich country, and in the less abstract sense of just having job opportunities fall in my lap. The small publisher that was my first gig turned out to be based walking distance from my house (or, more accurately, from my mum's house), and my fairly brief gig with the Dark Lord Murdoch came via a headhunter. I think I had to ask one or two magazines to let me write for them, but mainly they asked me.

You don't need this sort of implausible good fortune to make a Web site that makes a modest but live-on-able amount of money, but you do need a way to ride out the period of time while you make the site big and well-known enough for that income to build.

To do this, presuming you're not already wealthy or a kid living at home, you need to start the site as a hobby in parallel with a real job. Preferably the kind of real job that lets you sneakily work on your Web site while you're there, which can actually be done legitimately if you're a parking-station attendant or late-night petrol-station cashier or something, so a significant portion of your job description is "sit right there, and remain awake".

You also, of course, have to come up with some sort of idea for your site that can make money. The mass affiliate deals like Amazon or eBay are unlikely to be adequate, even if you do loathsome Sell Sell Sell stuff, as described in books that use the word "monetize". You need more direct deals with advertisers and retailers to make it work, as I did with Dan's Data and Aus PC Market. I made decent money when I reviewed Aus PC products; I made not much when I reviewed stuff from elsewhere. (And no, I didn't sell the free review product when I was done.)

Because of this, Dan's Data does not make me much money these days, because I burned out on reviewing computer gear years ago, and Aus PC gear reviews were my principal money source. If I were still writing about cases and CPU coolers and monitors all the time then Dan's Data would by itself still make me a passable living, but I just couldn't face another PSU or video card after a while, so now Dan's Data makes pocket-money only.

If you can start a site that covers some niche that (a) isn't already utterly saturated with high-quality journalism (or whatever you plan to do) already, and (b) lets you hook up with a business or three for mutual benefit, you absolutely can still start and run a Web site for a living.

Hell, if you're good enough you can even make adequate money from plain old ads; that's how the superlative Rock, Paper, Shotgun works. They accept donations as well, but I only now discovered that, since their donation page is harder to find than my cunning combined e-mail/donation scheme.

(I think the excellence of Rock, Paper, Shotgun and numerous other big game-review sites qualifies that market as "utterly saturated with high-quality journalism"; I wouldn't pin too much hope on a new game-review site making its owner much money these days. If you write good stuff, though, you can at least count on sites like Rock, Paper, Shotgun and news sites like the extremely venerable Blue's News to link to you fairly often. Starting a site that competes with Blue's News, Slashdot and other news sites that've all been taking body blows just from direct review-site RSS feeds a while ago, then Digg and now Reddit is, needless to say, not likely to be an express train to boundless wealth.)

That ain't workin', that's the way you do it!

I received this yesterday:

From: [redacted]
Date: Sat, 6 Oct 2012 20:49:46 -0700 (PDT)
Subject: Dan's Data Page

Hello Dan,

I came across your website,, this evening. I have been considering doing something like this for awhile. I was wondering if you would be willing to share with me how succesful it has been? I am trying to save enough money and invest it so I can live off of dividend payouts. My goal is to be able to be home with my family as much as possible. I have a target of atleast 100k and have managed to to save about 40k on my own thus far. It will takes years however to complete my goal on my own. I need a way to boost my savings. Please help.

[name redacted]

I always wonder how these people come to e-mail me. I've had two this week. I suppose they find my contact-and-donation page, which is titled "Give Dan Money For No Very Good Reason!", and... that's all they read, before clicking the e-mail link.

Because otherwise, they'd notice that people occasionally drop a buck or two in my tip jar because I, you know, wrote a load of stuff, on a very wide range of subjects.

Perhaps I've got this guy all wrong, and what he wants to do is start a Web site and slog away at it for a decade and make money that way. I suspect, however, that he, like the others who more clearly express their desire that I share my money-making secrets, just reckons I must be some kind of expert Internet panhandler. The contact/donation page scores really high in a Google search for "give me money"; I think a search like that is usually where these people come from.

When one of these correspondents seems to have two brain cells to rub together I direct them to my reply to this letter, in which I explain why people occasionally give me money. But all you really need to do is actually read the donation page, on which can be found subtle hints that it is not quite the only page on Dan's Data.

It'd make more sense if these e-mails were widely-copied scattershot spam, but they never seem to be. (Or, at least, Googling a string from them never turns up copies elsewhere.) Even the ones that include a sob story and ask me to send some of my presumed riches to them on account of how their son only has a burlap sack full of leaves for a body, or something, appear to have been typed in by an actual human and sent to only a few recipients, and quite possibly only to me.

I suppose sending spur-of-the-moment e-mails to someone who might know about getting strangers to send you money for nothing is a better wealth-generation strategy than just visualising money really hard and waiting for your Ultra Advanced Psychotronic Money Magnet to kick in.

I think you'd probably do better by just sending out PayPal money requests at random, though.

(The people I get PayPal money requests from almost certainly find me via the contact/donation page, too. Only seldom does someone really put in some effort.)

Zero to Kafka in five minutes, or no money back

OK, whippin' up the ol' Business Activity Statement for the first quarter of this year, tum te tum, run the special government BAS-management software and... it tells me I'd better renew my AUSkey certificate before it expires at the end of July.

Bit of an early warning, but OK, fair enough, off we go to the AUSkey site, which I leave open for a while as I enter stuff for my next BAS in the other special software that apparently a few other governments inflict upon their populace. (As is traditional with such things, this program likes to pop up dialog boxes telling you to enter a date for something, when you've entered data in some other field first, and then click on the date field intending to do the thing it is haughtily preventing you from doing until you click "OK".)

A few minutes later, I come back to the AUSkey site and click "login", whereupon it tells me my session has timed out and I have to go back to the home page, which is exactly where I already was.

What session? I don't have a session yet! I haven't logged in!

OK, argh, whatever, I log in again and it tells me I don't have the special AUSkey software which I thought I had but OK, again whatever, click the thing to download the software and... back I go to the home page again.

Go through that loop again until I realise that the site is attempting to tell me via mental telepathy that it does not support Chrome. Try Firefox instead, which to the government's credit does actually work and lo, now I do have the software that I installed whenever I went through this palaver the last time, and it doesn't even seem to need 283 updates since I last used it!

Righto, off we go, let's renew our certificate...

Hang on - there doesn't seem to be an option to do that anywhere.

Gee, could that perhaps be because the AUSkey does not, in fact, ever actually expire?

Why yes, that is the case.

Did the other program really tell me to update my AUSkey?

I quit it and run it again, and it doesn't say shit this time. I could have sworn it said I had to renew my AUSkey certificate but... now I... I just don't know.

You know, part of the reason why I wish Australia didn't have any submarines is that I'm not sure anyone's ever clearly explained what purpose they're expected to serve. (They apparently performed quite well in war games against the USA, which I'm sure will make all the difference if we decide to go to war with America. Or, marginally less crazily, with China, whose attack subs only outnumber ours 59 to six, not counting their entirely insignificant five nuclear ballistic missile submarines.)

Most of the reason, though, is that I don't think an institution that can create a system like this should be be allowed anywhere near explosives.

Catches fire, would buy again, AAA+++!

I bought a couple of sets of red LED lights on eBay; two ten-metre 100-LED strings for $US15.96 delivered.

You know the ones. Little lightweight controller box that always starts in cycle-through-all-modes mode, with a button that has to be pressed seven, or is it eight, times to get the darn things to just stay on constantly (or as close to it as the flickery PWM controller can manage).

Generally these cheap lights seem great. I've been very pleased with the others I've bought in the past, most recently the 220V-rated multicoloured ones from this seller, which seem to work very nicely from Australian power.

So I bought some red ones, alleged to work from 110 to 220 volts, from this other seller.

I plugged these new ones in while holding the wound-up lights in my hand, just to see if they worked at all, and they seemed OK.

And then, there was a pain.

In my hand.

A... burning pain, restricted to a few very small spots.

This puzzled me.

I adjusted my grip to avoid the ouchy spots, and observed a few thin trails of smoke rising from the wound-up lights.

I unplugged them.

I tried the other set.

Same deal.

These sorts of LED lights are configured as several long series strings, with a single inline current-limiting resistor (which, being one resistor at the start of a long series string of LEDs, probably doesn't actually limit current very well at all) in series with the first LED in each string.

[UPDATE: Now that I'm peeling one of the lights apart, it's apparent that they've actually got resistors on several of the LEDs early in each string. Here's a great analysis of these things and how to stop them flashing and flickering, forever.]

These resistors were getting very hot, very fast, and raising smoke from the clear PVC insulation over them.

Seizing the opportunity to use my variac and its delightfully mad-scientist-ish giant knob, I tried feeding the lights 110V instead of Australia's nominal-230V mains.

Now, they worked fine. The resistors got a bit warm, but not unduly so.

Fault located, then.

Next, like a damn fool, I told the seller that they were selling devices that were a fire hazard in 200V+ countries, and they should probably stop doing that, and could I have my money back, please?

Anybody who's ever filed an eBay/PayPal dispute over a defective item of low dollar value sold by some dude in China knows what happened next.

I opened a Dispute, I asked for a refund, they told me to get lost. I escalated the Dispute to a Claim, and eBay/PayPal in their wisdom told me to send the items back to the seller via registered mail to get my refund, which would of course be five bucks less than it'd cost to send the goods back.

(And if the seller decided to tell eBay that what I'd sent them was a box of newspaper, I probably wouldn't even get that.)

Perhaps if I'd lied and said the goods never showed up at all, I might have had a chance. Since I tried to warn the seller about maybe setting their customers' houses on fire, though, I got to pay the price.

Which is not in itself a big deal, of course, besides THE PRINCIPLE OF THE THING GRRR. It's not a dead loss, either; I can always chop the LED strings off the controller box and run them from some appropriate non-flickery DC power supply. This is not very difficult to do, and involves a lot less soldering than building an LED array used to.

I feel such a tit, though. Every time, I go through this idiotic routine, like Charlie Brown with Lucy's damn football.

Sometimes there's a bit of variety, like when I was trying to get a refund for an item described as new which turned out to be used, and the Hong Kong seller seemed to sincerely believe that "but if I give you a refund, I will lose money!" was an ironclad reason why he need not do so.

(Eventually he tried "OK, we'll give you a few bucks back, provided you lie in your feedback and say there wasn't a problem.")

I love the PayPal replies, too. You've proved that sending the item back will cost more than the refund? Well, now apparently it's a "judgement call" whether you should do so!

And then, "We know situations like this can be difficult and appreciate your patience and cooperation as we work toward resolution."

I really wish eBay/PayPal would be realistic in these exchanges and just say "hey, it's a flea market, almost always it works OK, but you got ripped off this time, it happens". Instead, just to twist the knife, when you give up and Cancel a PayPal claim, " agree that this complaint has been resolved to your satisfaction"!

(The only alternative is to wait until the clock runs out, whereupon PayPal tell you that the lack of resolution of your complaint is entirely due to your tardiness.)

So, in summation: EBay/PayPal aren't getting any better about this stuff.

And, if you're in Australia and want cheap twinkly LED lights in many colours, try these.

And don't buy stuff from this dickhead.

UPDATE: Lo, a message has arrived from the dickhead him or her self!

I'm sorry for that that our product make you no happy,
anyway, can you help to revise the feedback to positive and we'll refund

Yeaahhh... no. Product still fire hazard. Bad seller! Bad!

Please at least pay us for the postage

A reader writes:

I don't know if this really qualifies as a science question, but: How do bills for zero dollars and zero cents get sent to people?

I'm guessing that it's something like the process you explained that leads to crazy negative numbers in installer disk-space calculations - maybe the system's screwed up and thinks you owe negative money (or you really do, because you overpaid), but there's a sanity check that rounds negative numbers to zero, and then fails to stop a bill being sent anyway.

Am I right?


Exactly this sort of thing has been referred to since time immemorial, or at least since the early days of office automation... a "computer error".

Which it isn't, of course. Once in a very long while a cosmic-ray strike or failing power supply or actual hardware defect does really cause a computer to make an error, but the overwhelming majority of "computer errors" are actually programmer errors. The computer is doing exactly what it was told to do, whether that's sending a bill for $0.00, or charging a startled pensioner for a hundred million kilowatt-hours of electricity, or falsely saying an e-mail came from

One programming error that can lead to a zero dollar bill - and, in due course, to second requests and final demands and then the attention of lawyers, if only because the lawyers are happy to round the ten seconds needed to recognise the mistake up to about ten billable hours - is using the wrong kind of variable.

In the real world, money comes in dollars and cents, pounds and pennies, rupees and extremely un-valuable paise, and so on. There's no such thing as a fraction-of-a-cent coin.

In computers, everything can be chopped up into arbitrarily small pieces, if necessary. There may be some good reason to do this for certain calculations applied to currency amounts (though I can't think of one), but if you do, before any of the results get turned into actual money being paid into accounts or demanded from clients, the numbers should be rounded into an integer number of cents (or whatever other currency unit's being dealt with).

(This process can be exploited, in the classic "salami slicing" scam where a great many fractions of a cent are sneakily diverted to the scammer's account by, for instance, always rounding down, even when the fraction being rounded is larger than 0.5.)

If a programmer uses, say, a single-precision floating-point variable to hold a monetary value, it's easy for the limited precision of the variable to, when mathematical operations are performed on it, end up at 0.9999 (probably with a few more digits) instead of 1, or 0.0001 instead of zero. In the latter case, a system that doesn't round off the imprecise value to what it should be, and subsequently starts the billing process on any account where the amount owing is greater than zero, will send idiotic zero-bills to customers.

(Or the only slightly less stupid version, a bill for an amount less than what the company paid to post the bill to you.)

There are many other ways for this to happen, though, thanks to the many ways in which programmers can make a mistake. A billing system could, for instance, fail to notice previous overpayment, decide to send a bill, then apply the overpayment to the account balance and send a bill for the net amount, which could be zero or negative. Or it could decide to send a bill to a customer who legitimately owes money, and then accidentally print the amount that some other, fully paid-up, customer owes on the bill.

Just because a company's got thousands of employees and annual turnover greater than the whole economy of some nations does not preclude this from happening. As Daily WTF readers know, staggeringly expensive "enterprise" software can be very, very badly written.

Psycho Science is a regular feature here. Ask me your science questions, and I'll answer them. Probably.

And then commenters will, I hope, correct at least the most obvious flaws in my answer.

You have money you didn't know about! Give us some of it!

I love it when I don't have to go looking for an interestingly fishy business proposal, because some obliging organisation mails it to me.

(It's even better than unsolicited crank e-mail.)

Fishy letter

Strictly speaking, this one wasn't actually mailed to me, but to my partner Anne. It wasn't precisely aimed at her, either; they had our old address right, but if the recipient's name had been Norma Jeane Baker, the letter would have been addressed to Jeane Norma Baker.

So anyway, it's from an outfit called "CollectionPoint", and they're pleased to tell Anne that there's $AU887.50 waiting for her in an undisclosed location. Apparently CollectionPoint do debt recovery too, for a fee of 25% plus GST. They don't quote a fee for this other kind of money recovery, but I think it's safe to say it's not small.

We're not exactly rolling in dough at the moment, so a forgotten nest-egg could be quite handy.

(Do send me some money if you feel like it. We're hardly on the bread-line, though; I assure you that the lights will stay on, the cats will still get their little tins of fancy fish and the freeloading cockatoos will get their seed without your kind assistance.)

The questions that immediately occurred to me were, of course, "does this money actually exist?", and "is this outfit charging a fee for something you can do quite easily yourself?"

The answers to these questions are surprising and unsurprising, respectively.

"Unclaimed money" has been a scam-artist favourite for a long, long time. Unexpected inheritances. Prizes in lotteries you never even entered. A permutation in which the money may not actually strictly speaking be yours, but a morally upstanding person says you can still get hold of it, for a price. Some sort of purported government involvement. The list goes on.

The unclaimed-money business has even spawned meta-scams, in which the sucker pays for an information pack or franchise opportunity or something so they can start a work-at-home business finding unclaimed judicial judgements, or whatever, and creaming off a fat commission.

But CollectionPoint actually are telling us about money we really can claim. We'll claim it as soon as we can make a big enough pile of ID documents.

CollectionPoint are also, however, offering to take people's money to help them do something that is not actually difficult to do - or at least not significantly more difficult to do - by yourself.

The Australian government has a site called "Moneysmart" that'll point you at various unclaimed-money searches. Anne found the money CollectionPoint are talking about via the NSW Office of State Revenue site. Which is presumably the same way CollectionPoint found it.

So CollectionPoint do provide a helpful service. They alert you to the existence of money you probably can actually collect. And then you can throw the CollectionPoint letter away and go and collect your money the free way. CollectionPoint do not appear to be breaking any laws.

Well, they're not breaking any laws right now, anyway. The Australian Government's Department of Veterans' Affairs are happy to list CollectionPoint on their scam information page - apparently CollectionPoint sent letters to war widows claiming to be acting on behalf of that Department. And it's not hard to find other people talking about CollectionPoint in not-entirely-complimentary terms.

CollectionPoint come off pretty well in this blog post, for instance, until several allegedly separate people show up in the comments, all loudly defending CollectionPoint and all suffering from a suspiciously similar inability to construct a sentence, or in many cases even a word.

CollectionPoint also score themselves a mention in this Age article; apparently CollectionPoint have sent out follow-up letters implying - but not exactly actionably saying - that if you don't use their services, you'll miss out on the money altogether.

A commenter here says that after a CollectionPoint letter put him onto some money he could claim, and he claimed it himself without using CollectionPoint's services, CollectionPoint sent him a bill.

This bloke says CollectionPoint offered to collect $500 owing to him for a mere $160 - a 32% fee. Even Today Tonight doesn't like them.

Oh, and according (PDF) to the Consumer Action Law Centre, CollectionPoint charged a 25% fee for recovering some unclaimed superannuation money for an elderly client after he provided them with the identifying information he could have used to get the money back for free. But then CollectionPoint jacked up the 25% fee by adding another 10% GST charge (so 27.5%, altogether). The Consumer Action Law Centre took the case to court, and (another PDF) the Victorian Civil Administration Tribunal decided that CollectionPoint were indeed gouging their client, and reduced the fee payable to CollectionPoint by 45%.

The funny part, though, was that in response to this lawsuit CollectionPoint filed their own, in the same court, against the Consumer Action Law Centre's lawyers. They alleged "misleading and deceptive conduct" and an obscure kind of defamation, "injurious falsehood", which is becoming less obscure after recent reforms to defamation law in Australia.

In my non-lawyerly estimation, I think the result of this counter-suit can fairly be described as "widespread puzzlement".

So anyway, we're getting our eight hundred and something bucks.

CollectionPoint won't see a penny of it.