Why data on various elements is wrong
I got an email recently from a reader of my periodictable.com website asking why some subtle data on a couple of obscure radioactive isotopes seemed to be in conflict with other sources of data he had consulted. I responded by explaining why lots of the information you find listed in tables of data about the elements is inconsistent between sources. The short answer is the no one actually cares enough about it to fix errors.
A classic example of this is the question of which is the most dense of all the elements. You might think something as basic as that would have a clear and definite answer, but it doesn't. Over the past hundred-odd years, the title has switched several times between osmium and iridium.
There are three principle reasons for this. First, to be fair, their densities are very similar.
Second, it is very difficult to measure the density of a these very high-melting-point metals. The "density of an element" is widely agreed to mean the density of a perfect crystal of perfectly pure material (with its natural isotopic distribution) at standard temperature and pressure. Of course you can never create such an idealized perfect crystal, but you can try to get close. Or, better yet, you can use indirect means, such as measuring the lattice spacing with x-ray crystallography and combining that with the atomic weight. All these methods are difficult to execute, and error-prone. Thus they are expensive and time-consuming to carry out to a high standard.
Which brings us to the third impediment to getting a straight answer: It doesn't actually matter, so who's going to pay to do the hard work? No one in practice ever works with these two elements in any forms remotely close to their theoretical density, and even if they did, it's not likely to make a difference to any practical application what the precise density is.
It's also of very little theoretical interest. No great insights into the nature of matter will come from teasing out another decimal point in the density of osmium. It's just a number. Sure, it's interesting that there are certain zones of the periodic table that are much denser than others: From this fact we learn many things. But not from the fourth decimal place of any particular value.
This lack of importance applies to a great many of the physical properties of the pure elements. Sure, there are some (king among them silicon, but others as well), that are used in hyper-pure form in important applications. Their values are know with great precision, and sometimes the most accurate values are actually kept as trade secrets by companies that went to great expense to measure them more accurately than the competition. But this is the exception to the general rule that few elements have important applications in their pure form, and thus few values associated with pure elements are of practical significance.
So the situation with osmium and iridium is simply this: No one in the past hundred years has actually cared enough to settle definitively, once and for all, which the most dense. Maybe the current values are right, maybe not. Who cares. Not me, that's for sure. Notice how I haven't mentioned which is currently thought to be the most dense? That's because I don't remember, and can't even be bothered to look it up for my own blog post.