Of the different preparations of arsenic, it may be said in general terms, that those are most active which are most soluble. In conformity with what appears to be a general law in toxicology, the metal itself is inert. It is difficult to put this fairly to the test, because it is not easy to pulverize the metal without a sufficient quantity being oxidated to cause poisonous effects. Bayen and Deyeux, however, found that a drachm carefully prepared might be given in fragments to dogs without injuring them; and they once gave a cat half an ounce without any other consequence than temporary loss of flesh.[578] Its alloys are also inert. The same experimentalists found it inactive when combined with tin; and Renault likewise found it inactive when united with sulphur and iron in the ore mispickel, or arsenical pyrites.[579] It is probable that all the other preparations of arsenic are more or less deleterious. A difference of opinion prevails as to the power of the sulphurets. Various statements have been published on the subject. But it may be sufficient to observe, that in consequence of the poisonous properties of the sulphurets having been imputed to the oxide, with which they are often adulterated,—Professor Orfila made some experiments with native orpiment and realgar, and with the sulphuret procured by sulphuretted-hydrogen gas (which are all pure sulphurets); and he found that in doses varying from 40 to 70 grains they all caused death in two, three, or six days, whether they were applied to a wound, or introduced into the stomach.[580] It may appear at first view singular that the sulphurets, being insoluble, should be poisonous; but the apparent anomaly vanishes on considering the experiments of M. Decourdemanche formerly noticed; which prove that in animal fluids the sulphurets are rapidly changed into the oxide (see p. 225). The sulphurets, however, are much less active than the preparations in which the metal exists already oxidated. Yet in sufficient doses they will prove rapidly fatal. In the Acta Germanica there is the case of a woman who was killed in a few hours by realgar, mixed by her step-daughter in red cabbage soup.[581] The common artificial orpiment procured by sublimation is very active, in consequence of the oxide mixed with it. Renault found three grains killed a dog in nine hours.[582] Among the less active preparations of arsenic may also be enumerated such of the arsenites and arseniates as are not soluble in water. They have not indeed been actually tried. But there can be little doubt that they will prove poisonous; because, though insoluble in water, they are probably somewhat soluble in the animal juices. We may infer from their sparing solubility, even in these menstrua, that they will be less active than the preparations now to be mentioned, which are more soluble.