AND ZOOGEOGRAPHY OF INDIAN OLIGOOH^TA. 115 



the transition between these two genera is gi-aclua], and in some 

 cases the brandling of tlie centi'al canal of the prostate or its 

 absence can only be determined by microscopic examination. 



In the tliird line which starts from Plutellus the initial change 

 is the modification of the prostates; Wooclwardia, having thus 

 i-aeemose prostates, lumbricine setae, and meganephridia., cannot 

 be placed on either of the other lines, since in them either 

 the setse or the nephridia are modified from the start. From 

 Woodwardia is probably to be derived Comarodo^ilus, in which 

 the nephridia in front of the clitellum, but only these, are broken 

 up, the gizzard has become vestigial, and the originally paired 

 spermathecal pores have fused in the middle line. 



The genus Spenceriella lias the primitive form of prostate, but 

 is micronephridial, and has the perichaetine ari'angement of 

 setae ; it is probably to be derived from Alegascolides by mul- 

 tiplication of the setse. It could however equally come from 

 Bifporocliceta by the breaking up of the nephridia. 



A group of small genera are characterized by the reduplication 

 of the gizzard. Digaster and Didymogaster have two gizzards, 

 and ai^e distinguished from each other by the number and 

 position of the spermathecre ; Perissogaster has three gizzards 

 situated anteriorly, as in the two former species. The condition 

 of the other systems indicates that these are all to be derived from 

 Notoscolex. Plionogaster, in which there are several gizzards 

 more posteriorly situated, at the beginning of the intestine, is to 

 be considered as originating from Megascolex. 



Finally Pontodrilus is to be mentioned. The majority of 

 species are littoral in habitat; one is terrestrial, and one is limnic. 

 It is derived directly from Plutellus ; the gizzard has become 

 vestigial, and nephridia are absent from the first twelve or 

 fourteen segments. 



Attention may here be drawn to two points. The first is 

 that the genera of this subfamily can be arranged in phylo- 

 genetic order. We know which characters are primary, which 

 secondary — and therefore we know which forms must have come 

 first in evolution ; in addition, we have in several ca,ses inter- 

 mediate forms (betweeii Diplotrema and riutellus, which are 

 united by Michaelsen (14 a); between Megascolides and Noto- 

 scolex; between Kotoscolex ^iXid Megascolex, which Michaelsen also 

 merges (16); and between Dipo7'Ochceta and Perionyx, which again 

 a,re united by Michaelsen (16)). The whole tree is still before 

 us, and all stages in the evolution of the subfamily are there for 

 detailed examination. While there is room for doubt in some 

 details, the main outline will probably stand firm. 



The second point is that evolution has proceeded along a few 

 definite lines ; the essential changes are confined to a few 

 systems, and follow a definite direction in each case. We have 

 the change in the arrangement of the setae, from the lumbricine 

 to the perichaetine; the change in the nephridial system. 



