220 SUMMARY OF CURBENT RESEARCHES RELATING TO 



end-organs of the larval lateral line are finally transformed into 

 glands ; rather do they seem reduced to ordinary epidermic cells. 

 Most characteristic of the adult epidermis are its " flask-cells," so 

 termed from their peculiar figure. These are developed periodi- 

 cally by the transformation of cells of the stratum mucosum. They 

 appear contemporaneously with each new stratum corneum ; they do 

 not seem themselves capable of division. " They belong morpholo- 

 gically and physiologically to the stratum corneum, and have, like 

 this, a purely mechanical function." Various interpretations touching 

 the genesis and office of these cells, which occur also in frogs and 

 tritons, have been suggested since Eudneff described them (in 1865). 

 This observer found them in the frog. Langerhans afterwards 

 detected them in the salamander, and suspected a connection between 

 them and Leydig's cells. 



Good examples of indirect cell-division, confirming Flemming's 

 views, were found by Pfitzner in the course of these investigations. 

 (See his figures.) 



The well-known colours of the salamander are not due to the 

 pigment of the corium, but to that of the epidermis, which is partly 

 diffuse, partly contained in special cells (chromatophores). The 

 origin of these cells is not known ; are they truly epidermic, or have 

 they come from the corium ? There is little difference, save as to 

 degree, between the pigmentation of the adult and of the larva. Parts 

 which are faintly coloui-ed in the former, as the under surface of the 

 head, are usually without pigment in the latter. By confining larvse, 

 from their birth, in shallow porcelain vessels kept in the dark, as 

 suggested to our author by Dr. Boas, less pigment becomes formed ; 

 such specimens are obviously favourable for microscopic study. The 

 pigment-granules of the stratum corneum generally form a layer just 

 within the cuticular border ; those of the stratum mucosum more 

 affect the upper (outer) halves of the cells in the larva than in the 

 adult. The cell-nuclei and intercellular spaces never contained 

 loose pigment-granules. Leydig's cells, also, showed no pigment. 



The cornea, which is very unlike that of the frog, takes no share 

 in the successive moults, but retains throughout life the characteristic 

 structure of the epidermis of the newly-born larva. It is well sup- 

 plied with intercalary cells, but has no flask-cells or cells of Leydig. 

 The system of intercellular spaces opens freely upon its surface. 



Of all parts of the body the epidermis is most immediately, con- 

 stantly, and extensively related to the environment. Its passive as 

 well as active functions are plainly, therefore, of the first importance. 

 We are thus led to note the contrasting structures which it displays 

 among fishes and larval amphibious vertebrates, on the one hand, and 

 typical terrestrial vertebrates on the other. The stratum corneum 

 in the former is represented, functionally at least, by the cuticular 

 border of the outer tier of cells. There is an unmistakable resem- 

 blance between the soft tegumentary and alimentary epithelia ; both, 

 indeed, are exposed to a liquid medium. In the higher vertebrates 

 we find a polyderic stratum corneum. The adult salamander, like 

 other members of its class, here occupies an intermediate position 



