ZOOLOGY AND BOTANTj JHCEOSCOPYj ETC. 659 



epidermal cells ; the nutrient material for this activity is clearly 

 afforded by the fused mass of yolk-cells, the nuclei of which may be 

 observed to be undergoing retrogressive metamorphosis. No evidence, 

 however, could be obtained as to the direct conversion of the yolk-cells 

 into endodermal elements ; we seem rather to have here to do with a 

 remarkable substitution of organs. The yolk-cells appear to be vica- 

 rious endodermal cells, which have undertaken their duty in conse- 

 quence of a shortening of the process of development. Primitively a 

 true primary endoderm must have been developed, and then the yolk- 

 cells served as nothing more than nutrient material. Indications of 

 the remains of a primary endoderm are to be sought for in a small 

 group of cells which lies below the pharynx, the early development 

 of which organ clearly stands in relation to the important part played 

 by the yolk-cells. It has not till yet been known that a rudimentary 

 organ could become differentiated before the formation of the germinal 

 layers or the body-wall ; here, however, we have an example of a 

 pharynx being developed before the definite ditferentiation of the 

 ectoderm and mesoderm. This being so, we must not wonder if the 

 central nervous system of the fresh-water Triclades shall be shown to 

 be formed from the mesoderm. It would hardly be so remarkable an 

 example of cenogeny. 



Eemarkable as are the embryonic adaptations presented by these 

 forms, they are not without their like in the animal kingdom ; thus, 

 in the Tunicata, the atrial cavity may be developed in very different 

 ways, and by modes which can hardly be supposed to be the primitive 

 ones. The mesodermal origin of the rudiment of the cerebrum has 

 been noted in some Mollusea, and there are, in fine, quite a series of 

 embryonic adaptations which affect most intimately the differentiations 

 of the germinal layers. Indeed the Tunicata, where, as Salensky has 

 shown, the blastomeres of the embryos of Saljja disappear to be re- 

 placed by follicular cells, afford even better than Planaria folychroa, 

 the most striking examples of cenogeny. 



Echinodermata. 



Physiology of EcMnodermata,* — Dr. G. J. Romanes has already 



shovni I that the ocelli at the end of the rays in starfish, and occupy- 

 ing the homologous position in Echini, perform a visual function — 

 inducing the animal to seek the light so long as the ocelli are intact, 

 and the animals ceasing to be affected by light when the ocelli are 

 removed. He has now tried whether these organs might not have an 

 olfactory as well as a visual function. Having procured some fi-esh 

 starfish, he dropped little pieces of limpet and crab in their vicinity. 

 None of them, however, approached the food. Supposing it possible 

 that the starfish, having been freshly caught, might not be in want of 

 food, he left them in a tank for a couple of days, and then repeated the 

 experiment. The result was now quite different, for the starfish began 

 actively to crawl in the direction of the food. Selecting, therefore, 



* Joum. Linn. Soc. Lond. (Zool.) svii. (1883) pp. 131-7. 

 t Croonian Lecture, Phil. Trans., 1881, pp. 829-85. 



