paper referred to was read I learned that Mr. Park, late Lecturer at the Thames School of Mines, had made a series of valuable experiments on the subject, in which he used the gravimetric method for determining whether there was not a dissolution of gold by aqueous solution of this gas.

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

The results of these experiments do not confirm the correctness of this statement of mine (that cyanogen is unable to dissolve gold), but, still, they clearly show that, at least, solutions of this gas do not dissolve this metal at all readily. Mr. Park informs us that he performed his experiments with a button of pure “parted” gold gently hammered to a coherent spongy mass of about ⅛in. in diameter, and weighing 0·340 grain. This button was placed in an aqueous solution of cyanogen and the containing vessel loosely covered. Upon again weighing this button, at intervals of twenty-four hours, he found there was generally a loss of about 1/330 of its weight at every weighing—that, in fact, about 1/1000gr of gold dissolved per diem.

This does not appear to be a great loss; still, it is very much more than I should have anticipated, but, knowing that the most recent works on chemistry to hand in the colony decide that cyanogen in water alone does not decompose to substances solvent of gold, this scientist could not support my contention as he desired to do.

It was this unsatisfactory state of the case that induced me to make further researches in the matter. Now, as you are aware, cyanogen is a substance that in the presence of even minute traces of ammonia ox potash is decomposed to form alkaline cyanides which are solvent of gold, and when once this action starts it proceeds with ever-increasing rapidity. The atmosphere of a laboratory in full operation is frequently alkaline; the vessels used for receptacles in chemical work are capable of yielding alkaline matter to cyanide solutions. For these reasons any experimental results obtained in a laboratory are likely to be misleading.

It is evident, therefore, that any method which requires considerable periods of time, such as the gravimetric method does, is not well adapted for this kind of research; one is required that will speedily give reliable results. Eschewing, therefore, the use of even the hypothetical just balance as an abomination in this case, I adhered to my old method, which is that of testing by sight alone whether any loss of gold does occur by the action of free cyanogen. For this I merely replaced the gold leaf of my former experiments by gold paper, which is a Swedish filter-paper, in which gold has been chemically precipitated in a very finely-divided state. A sample of this test-paper is tabled here for exhibition, and the red tint of its gold is easily perceptible in this paper,

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

1/10in. square, by contrasting it with the same kind of paper that has not been so treated. A few short statements showing the extreme tenuity of the gold in this paper may be interesting.

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

A square inch of the paper contains 1/10000gr., and 1/10in. square contains 1/1000000gr. of gold. Were the gold in this paper agglomerated to a film having a like area with that of the containing paper that film would be only 1/50000000in. thick—that is, 250 of these would be the thickness of gold leaf. In the paper itself (being, as it is, 1/120in.thick) this film (of the 1/50000000in.) is broken up to occupy a volume 400,000 times that which it occupies in the form of a film.^*

It follows, therefore, that the gold in this paper, volume for volume, only weighs half as much as hydrogen gas.

Broken up in this manner in the test-paper before you, it is in very truth fine gold—in fact, gold divided almost to its ultimate atom (if, indeed, atoms do exist)—gold in the cloud form, as it were, and therefore in the best condition that I know of for my purpose. Provided with a test so delicate as this is, we get results in an hour that, using the gravimetrical test for loss, would require several days, and so we avoid those errors that are apt to creep in and vitiate our results when long periods of time are required for experiments of this nature.

Placing then in a porcelain vessel a strong aqueous solution of cyanogen, along with a little of this gold test-paper, I closed the vessel down airtight, and on examining at periodic intervals I found that even after the expiration of six hours, corresponding to sixty-two days for gold leaf, there was no visible diminution of the colour of that test-paper. After this, however, the tint gradually faded, until in thirty hours it had quite disappeared. Thirty hours to dissolve the millionth of a grain of gold so finely divided as this gold was, shows that if cyanogen itself does dissolve gold it is only at an extremely low rate—at such a rate that ordinary gold leaf would require about one year to become entirely dissolved therein.

Now, this result is a very different one to those that I am faced with both by Professor Black and Mr. Park; still, while it is clearly shown that for gold-milling the gas cyanogen as a direct solvent is useless, it does show that there is an infinitesimal dissolution of gold either by cyanogen or its derivatives, and in the interest of exact science the question has to be decided which of these it is.

Now, the cyanogen I used, though very carefully prepared, had a slight acid reaction; it contained traces of ammonia, hydrocyanic, and hydrochloric acids, and this even when

to avoid producing them I used the protochloride in place of the bichloride of mercury; and I may further state that the solution of this cyanogen kept persistently acid, even when it was three weeks old and much of the gas decomposed; and, what is more singular still, the solution was far more solvent of gold at that time than just after it had been made. It was therefore apparent to me that it is not any degree of acidity in a cyanogen solution that does, as is now supposed, conserve this gas—it is not a mild acidity that is always effective for this purpose. So I increased the acidity of the cyanogen solution by adding to it a few drops of hydrochloric acid, and tried another gold test-paper therein, when I was unable to detect, even after the lapse of seven days, the slightest change of tint had been produced upon that test-paper. Hydrochloric acid had not any retarding effect upon the dissolution of gold in weak solutions of bromine.

For supplementary and confirmative evidence on this point I next passed cyanogen gas through a weak solution of nitrate of silver to wash out any ammonia, hydrocyanic and hydrochloric acids contained therein, and the purified gas was then allowed contact with the gold test-paper both as gas and as aqueous-solution of it, when I got results altogether confirmatory of those obtained in the previous experiment—that is, no perceivable effect was produced on the test-papers by seven days' contact.

The results of these various experiments, taken collectively, appear to be positively overwhelming in favour of the correctness of the assertion I made before the Society last month—that aqueous solutions of free cyanogen have not the least solvent power upon gold; consequently they support the old contention of mine cited in the former paper, and alluded to here—that cyanogen does not, as is now generally supposed, compare with the haloids, chlorine, bromine, and iodine, chemically—that, in fact, except that it appears to be a monad with these and a dozen or so more of the elements, it has no chemical relations to any of them.

I should inform you that the gold I used for the experiments here detailed, also for those for my former paper to the Society, was practically pure; at least, it only contained minute traces of copper. Argentiferous gold, of which class most or all our native gold is, would, of course, if possible, be still less amenable to solutions of cyanogen than the gold I used, and for the reason that argentic cyanide, if formed at all, would always remain as a product quite insoluble in such cyanogen solutions

That cyanogen would have very little tendency to form by its decomposition solvents for gold as used upon quartz, &c., at the gold-mines appears to me absolutely certain, as both air