58 Records of the Indian Museum. [Voi,. XXIV, 



cavities are considerably smaller and do not show any proto- 

 plasmic elements within their limits. The so-called columnar 

 epithelium is not well-marked. The underlying structure corre- 

 sponds to that described for the vertical section. 



Having described the structure met with in the different forms 

 selected above, it will be advantageous to consider the lines along 

 which evolution has taken place from the simplest to the most 

 complicated structure. In the simplest type of adhesive appa- 

 ratus the skin is thickened and its outer epithelial layer of cells is 

 modified into curved spines; the gland cells and the " clavate 

 cells" of normal skin-tissue totally disappear. Specialization 

 proceeds along two lines, firstly there is an increasing tendency at 

 every step towards vacuolation of the superficial epithelial cells 

 and this ultimately results in the formation of definite cavities, 

 and secondly the cell-walls of epidermal cells become indistin- 

 guishable. The final stage is reached in Glyptolhorax sp. where 

 the whole of the epidermal tissue appears as a syncytium. The 

 adhesive apparatus of the Silurid fishes is distinguished from that 

 of the Cyprinid forms by the fact that in the former the skin is 

 thickened and striated, whereas in the latter it is only thickened 

 and forms a plain cushion-like pad. I have not been able to 

 understand the true significance of increased vacuolation or of the 

 s)rncytium formation in the tissue of the adhesive apparatus. 



There is, however, no doubt as to the function of the epidermal 

 spines. All of them are curved in the same direction and probably 

 they point posterior^. As the fish sticks to a stone with the head 

 pointed up stream, and the current of water tends to move it 

 backwards, the spines when closely applied to a stone, fix it 

 securely by taking hold of the unevennesses of the rock. In 

 Silurid fishes the ridges are pulled outwards and a sort of vacuum 

 is produced in the grooves which helps the fishes in adhering to 

 rocks. 



So far as I know, the type of structure described above is not 

 met with in tissues of adhesion in any other group of animals. 

 Dahlgren and Kepner, 1 who have given a summary of the subject, 

 have not anywhere referred to a spinous structure. Even in a 

 longitudinal, vertical section through a small region of the grasp- 

 ing organ on the head of Remora, the characteristic structure 

 that I have described above apparently is not found. 



CONCLUSION. 



In conclusion I wish to refer briefly to the origin of the hill- 

 stream fauna and to the means of dispersal and propagation 

 adopted by it. In the following discussion I take up these points 

 one by one. 



Origin of the Hill-stream Fauna. — There are two possibilities, 

 firstly, that the forms now living in the mountain-rapids were once 



1 Dahlgren and Kepner, Principals of Animal Histology, pp. 409-417 

 (1908). 



