No. 32. — 1886.] scientific beseaeches. 299 



of a tail is discernible, and the embryos show clearly a short 

 tail with a well marked fin. These are the main features of 

 the development of Epicrium. There are many other points 

 of interest, but they come more into the scope of a specialist, 

 and could not be explained in a short paper and without the 

 aid of drawings, and they will have to be dealt with accord- 

 ingly at a later period. 



The chief results are briefly — 



(1) Epicrium (like Coecilia), on the basis of its anatomy, 



cannot be placed in any known group of Am- 

 phibians. 



(2) Its embryonic evolution passes through the same 



stages as the salamander as a larva. 



(3) In consequence, Epicrium, notwithstanding its very 



different appearance, has to be classed as the 

 nearest neighbour of the salamander. 

 The researches just referred to occupied us about eight 

 months at Peradeniya. We then procee ded to Nuwara Eliya 

 to begin studying the rain-worms of Ceylon, both from a 

 histological and a systematic point of view. The hills of 

 Ceylon are of special interest in possessing quite a gigantic 

 species of rain-worm, which attains a length of about four 

 feet, and a thickness of a big thumb. We hoped to find other 

 similar species, but we met only with smaller kinds 

 that were new, amongst which were a beautiful blue- 

 ringed one of rather large size, and many smaller ones of less 

 striking appearance. Ceylon being an island long since 

 separated from the continent, those animals living in the 

 soil, as they are, must be of a peculiar interest, and an exact 

 systematic examination of the rain-worms of Ceylon might 

 induce an investigation of those of India. The comparison 

 of the forms in both countries promises valuable results as to 

 the range of geographical distribution ; it might also furnish 

 a further basis for speculation about the changes of the con- 

 tinents and islands in the course of time. Now, it is true 

 that the rain-worm is very easily transferred in many ways ; 

 112-86 D 



