106 DR, J. STEPHENSON ON THE MORPHOLOGY, CLASSIFICATION, 



tufts, each apparently with a funnel ; in addition, in the hinder 

 segments there is on each side, near the ventral nerve cord, a 

 larger nephridium in the form of a fair]}' large rosette, with a 

 funnel in the preceding segment. Trigaster lankesteri has the 

 same arrangement, while Oclochcetus thurstonihas numerous micro- 

 nephridia throughout the body, without any trace of the larger 

 organs. The only morphological change, therefore, which is 

 necessary in order to evolve Eudichogaster from Trigaster is the 

 development of calciferous glands ; and indeed these are scarcely 

 present in one species of Eudichogaster. 



Michaelsen admits that the geographical facts do not at first 

 sight appear favourable to this view. As has been noted, 

 Eudichogaster is purely Indian, while Trigaster belongs to 

 Mexico and the West Indies. But the widespread occurrence of 

 Dichogaster, a descendant of Trigaster, in tropical Africa is 

 evidence, according to Michaelsen, of a former land-bridge across 

 the Atlantic; and Michaelsen supposes that either (1) Eudicho- 

 gaster originated from Trigaster in America, spread eastwards 

 across the land-bridge to Africa, colonized Africa or parts of it, 

 and then spread eastwards again across another land-bridge to 

 India (it is not found in Africa at the present day because it 

 has been extirpated there by the dominant Eudrilines and the 

 later evolved Dichogaster) ; or (2) that Trigaster itself spread 

 eastwards by the same bridges, and gave rise to Eudichogaster at 

 the eastern extremity of its range, i. e. in India, itself later being 

 extirpated in the middle portion of its range, ?'. e. in Africa, by 

 the Eudrilines and its own descendant Dichogaster. as before. 



In putting forward the view that the ancestor of Eudichogaster 

 is Octochoztus and not Trigaster, it may be admitted that, as 

 Michaelsen says, Eudichogaster and Trigaster are very much alike 

 morphologically. At the same time, I think we now possess 

 evidence of a much closer similarity between Eudichogaster and 

 Octochatus (or at least some species hitherto reckoned as Octo- 

 chcetus) than Michaelsen was aware of. The gap between the 

 two genera is bridged almost, if not quite, as completely as that 

 between Eudichogaster and Trigaster. The points to be discussed 

 are the gizzards, the calciferous glands, and the nephridia. 



I have described (11) in Octochcetus pallidus a commencing 

 doubling of the gizzard: — "The gizzard is barrel-shaped, in 

 segment vi.; the oesophagus is distinctly strengthened in seg- 

 ment v. also, where shining longitudinal muscular bands are seen. 

 This seems to be the beginning of a, double gizzard, such as seen 

 in Eudichogaster, Dichogaster and Trigaster : I do not, however, 

 suggest at present that any of these genera are derived from this 

 species, or indeed from the genus Octochcetus at all." 



I may here call attention to the relation of the septa to the 

 condition of duplicate gizzard. Octochcetus pallid%i,s is one of the 

 rather few species of the genus which retain all the septa in the 

 anterior part of the body (behind the level where they first 



