Professor Herdman subsequently conducted some experiments[158] with fishes, with the view of testing his theory that the shapes and colours of Nudibranchs serve the purpose either of protection or warning, and bear direct relation to the creature’s edibility. These experiments, on the whole, distinctly tended to confirm the theory. _Aeolis_ was evidently very nasty, and probably stung the mouths of the fishes who tried it. For the complete success of the theory, they ought to have let it severely alone, but the fish were evidently accustomed to make a dash at anything that was dropped into their tank. Another conspicuous mollusc, _Ancula cristata_, was introduced, Professor Herdman and his collaborator each commencing operations by eating a live specimen themselves. They found the taste pleasant, distinctly like that of an oyster. The fish, however, when the experiments were conducted under conditions which made the scene as much like ‘real life’ as possible, did not agree with Professor Herdman. The _Ancula_ crawled over various parts of the tank for several days untouched by the fish, who sometimes went close to them and looked at them, but never attempted to taste them. Experiments with species whose colours were protective, such as _Dendronotus_, were also conducted, and the decided edibility of these species was established, the fish competing eagerly for them, and tearing them rapidly to pieces. Mr. W. Garstang, of the Plymouth Laboratory of the Marine Biological Association, confirms[159] Professor Herdman’s views as to the shape and colour of Opisthobranchs. _Pleurobranchus membranaceus_ is known to secrete, on the surface of the body, an acid which reddens blue litmus paper. It is, therefore, no doubt distasteful to fish, which all abominate the taste of acids, and is conspicuously marked with red-brown and yellowish ‘warning’ colours. _Haminea_ and _Philine_, on the other hand, are good to eat, and consequently possess ‘protective’ coloration. _Runcina Hancocki_, which is of a brown colour, crawls over brown mud and weeds, but avoids green weeds, on whose surface it would appear conspicuous. _Elysia viridis_ varies its colour according to its habitat, being green when on green weeds, and dark olive, brown, or reddish brown, on pools among tufts of littoral algae. Green specimens of _Hermaea dendritica_ were kept in captivity, and placed in a dish with green and red sea-weeds. They were never observed crawling upon the red weed, upon which they would have been very conspicuous. _Archidoris flammea_ occurred on bright red sponges, to which its colour was so closely assimilated that Mr. Garstang at first quite overlooked it. _Goniodoris castanea_ was found under stones, feeding on compound Ascidians (_Botryllus_), which it sufficiently resembled to be very inconspicuous in that position.