60 MR. F. E. BEDDARD ON [Feb. 4, 



presence of a single pair of these organs in each segment is the 

 archaic condition, tlien Perichceta will be a modification of Acantho- 

 drilus ; but this view is confronted with two serious difficulties — (1) 

 the apparent specialization of a part of the nephridial network to 

 form a series of paired nephridia in' Megascolex armata and in 

 MegascoUdes australis (Spencer) will require explanation ; and (2) 

 the connection of the vasa deferentia with the atria ( = prostates) 

 will have to be regarded as having bteu derived from a condition in 

 which these organs are independent of each other {Acanthodrilus). 

 These structural peculiarities are capable of an intelligible explana- 

 tion if we assume that Deinodrilus is an intermediate stage in the 

 evolution of Acanthodrilus from Perichceta. 



The remarkable arrangement of the setae in certain Australian 

 Perichcefce, which I have in the present paper associated together in 

 the geuus Anisochceta, and the commencing reduction of the setae 

 in P. indica must be considered in relation to this question. It is 

 notevAorthy that in these cases it is the anterior segments only which 

 diifer from the posterior in the reduction of the setae. The forma- 

 tion of a "head" is also brought about by specialization in the 

 alimentary and excretory systems, and by a partial obliteration of 

 the coelom and loss of internal segmentation. These facts tend to 

 show that the reduction in the number of the setae is also secondary; 

 and this reduction is very general in Perichffitidae, though not any- 

 where so apparent as in Anisochceta. It is true that, as Perrier 

 first pointed out, the hindermost segments of Perichceta may also 

 show a reduction in the number of setae; but this fact may be in 

 accord with the views here advocated, inasmuch as the nephridial 

 system in MegascoUdes begins to be specialized in the posterior 

 region of the body. I would, however, rather insist upon the 

 increase in length produced by the addition of new segments at the 

 end of the body, and explain the few setae of these segments as due 

 to their recent formation and consequent imperfect development. 



If we were acquainted with a species of Lnmbricus in which the 

 anterior segments were provided with a larger number of setae than 

 ordinarily, it would certainly be set down to " cephalization ; " there 

 is therefore nothing unreasonable in regarding the converse change, 

 which actually occurs, as due to the same cause. These facts, there- 

 fore, are at least not contrary to the assumption that the " perichce- 

 tous " condition is the more primitive. Among the species of 

 Anisochceta which show the reduction to 8 setae per segment, some 

 have more segments modified in this way than others ; there is, 

 therefore, evidence of a gradual change in this direction which lends 

 more weight to the arguments here advanced than if all were 

 modified to exactly the same extent. In the latter case the facts 

 could be referred only to a modification affecting the " head" and 

 comparable for example to the loss of the setae in some of the first 

 few segments in Chcetogaster ; as it is the facts appear to point to 

 a gradually advancing reduction of the setae commencing in tlie most 

 modified region of the body. 



