_c._ Of two thornbacks, one, C was introduced into about an ounce of boiled water in contact with hydrogene, standing over mercury. The other, D, was introduced into well boiled water saturated with nitrous oxide, and standing in contact with it over mercury. C lived near thirteen minutes, and died without being previously much agitated. D was apparently motionless, after having the same affections as A in the last experiment, in sixteen minutes. At the end of this time he was taken out and introduced into common water. He soon began to move his gills, and in less than a quarter of an hour was so far recovered as to be able to swim. The last experiment was repeated on two smaller thornbacks; that in the aqueous solution of nitrous oxide lived near seventeen minutes, that in the water in contact with hydrogene, about fifteen and half. The experiments in Res. I. Div. 3, prove the difficulty, and indeed almost impossibility of driving from water by boiling, the whole of the atmospheric air held in solution by it; they likewise show that nitrous oxide by its strong affinity for water, is capable of expelling air from that fluid after no more can be procured from it by ebullition. Hence, if water saturated with nitrous oxide had no positive effects upon fishes; they ought to die in it much sooner than in water deprived of air by ebullition. From their living in it rather longer;[189] we may conclude, that they are destroyed not by privation of atmospheric air, but from some positive change effected in their blood by the gas. [189] A priori I expected that fishes, like amphibious animals would have been very quickly destroyed by the action of nitrous oxide. A long while ago, from observing that the gills of fish became rather of a lighter red during their death, in the atmosphere; I conjectured that the disease of which they died, was probably hyperoxygenation of the blood connected with highly increased animal heat. For not only is oxygene presented to their blood in much larger quantities in atmospheric air than in its aqueous solution; but likewise, to use common language, in a state in which it contains much more _latent heat_. Without however laying any stress on this supposition, I had the curiosity to try whether thornbacks would live longest in atmospheric air or nitrous oxide. In one experiment, they appeared to die in them nearly in the same time. In another, the fish in nitrous oxide lived nearly half as long again as that in atmospheric air. IX. _Effects of Nitrous Oxide on Insects._ The winged insects furnished with breathing holes, become motionless in nitrous oxide very speedily; being however possessed of a certain voluntary power over respiration, they sometimes recover, after having been exposed to it for some minutes, under the action of atmospheric air.