DEVELOPMENT AND METAMOKPHOSIS. 97 



whilst the latter takes only living prey. Nevertheless 

 the iutestiue in all tadpoles is exceedingly elongate, 

 being several times longer and of a greater calibre 

 than after the metamorphosis. Bataillon has ascer- 

 tained by direct experiment that the abbreviation of 

 the gut is not localised to any special region. Tad- 

 poles of Alijtes obstetricans having been anaesthetised 

 a few days before the egress of the fore limbs, the 

 abdominal cavity was cut opeu,and the intestine, after 

 having been slightly uncoiled, marked at regular dis- 

 tances by silk threads tied round it. The abdomen 

 was then sewn up, and the tadpoles, having recovered, 

 continued their evolution. At the close of the meta- 

 morphosis the gut was again examined, and it was 

 found that the threads, which were tied at intervals 

 of twenty millimetres, were then only seven or eight 

 millimetres apart. 



I have measured the uncoiled digestive canal in 

 fully developed tadpoles and in recently transformed 

 young of the three following species. It will be seen 

 from the measurements, in millimetres, how consider- 

 able the range of variation between different forms 

 may be, both in the larval and perfect conditions. 

 Thus in Bana esculenta the length of the intestine of 

 the tadpole is nearly ten times that of the perfect 

 form, whilst in Ahjtes obstetricans its length is only 

 as 4 : 1 ; Pelodytes punctatus, with 6:1, is inter- 

 mediate. This difference is correlated with the 

 diet, Bana esculenta being the most herbivorous, 

 and Ahjtes obstetricans, as noticed above, almost ex- 

 clusively carnivorous in the tadpole condition. 



Length of intestine in tadpole (i) and young (n) of — 



i. ir. 

 Ahjtes obstetricans (25 mm. body-length) 235 60 

 Pelodi/tes punctatus (20 ,, ., ) 194 32 



Bana esculenta (23 ,, ,, ) 472 45 



Before closing this chapter on Development, the 

 highly interesting fact remains to be alluded to that 



G 



