DESCRIPTIONS OF GENERA AND SPECIES 457 



Affinities of Cryptodrilidae. 



This family has perhaps the most intimate relations J with the Perichaetidae ; the 

 genus Megascolex forms an almost ideal link ; the fact that in the Perichaetidae 

 there are invariably more than eight setae per segment, is really the only dis- 

 tinguishing mark which enable them to be differentiated from the Cryptodrilidae ; 

 and it will be remembered that in the genus Megascolex the anterior segments may 

 have only eight setae apiece, while further back the setae become numerous. This 

 genus might with equal reason be assigned to either of the two families Perichaetidae 

 or Cryptodrilidae. It is not too much to say that there is hardly a single point of 

 structure in the Cryptodrilidae which is not also characteristic of the Perichaetidae, 

 and vice versa, that is to say of some one or more Perichaetous worms or Cryptodrilids. 

 At first sight there may be some objection to this statement. Many Cryptodrilids 

 have two gizzards, or even three ; while, with the exception of Pleionogaster, no 

 Perichaeta (in the wide sense) has more than a single gizzard ; and the gizzards of 

 Pleionogaster being at the posterior end of the oesophagus rather recall the similarly 

 placed gizzards of the Eudrilids, Hyperiodrilus and Heliodrilus, than Dickogaster or 

 Digaster. The genus Perichaeta, however (in the strict sense), has a gizzard which 

 occupies certainly two, if not three segments ; where there are two gizzards in the 

 Cryptodrilidae they lie in consecutive segments ; now there is not a wide interval 

 between the two kinds of modification ; and I should be disposed to compare the 

 single gizzard of Perichaeta occupying two to two separate gizzards in a Cryptodrild 

 occupying two consecutive segments, rather than to compare it with the single 

 gizzard of such a form as Eudrilus or Acanthodrilus. 



With other families of Oligochaeta there are not such plain relationships. The 

 Acanthodrilidae, for example, are separated by certain well-marked characters from 

 all Cryptodrilidae. 



MICHAELSEN (10) has certainly compared Dickogaster with Benhamia, giving a 

 tabular statement of the points of similarity and difference ; although these resem- 

 blances are undoubtedly numerous and striking, there is one difference which 

 absolutely distinguishes all Cryptodrilidae from all Acanthodrilids ; in every Acan- 

 thodrilid the male pores open on to the eighteenth segment ; and the spermiducal 



1 The intimate relationship which exists between all the families. of the Megascolicidae is, perhaps, 

 exemplified by the genus Hicroscdex of 8. America and its affinities to the Aamtltodrilus of that continent, and 

 by the Cryptodrilids and Periehaetida of Australia. The sperm-sacs of the American Microscolex and 

 Acanthodrilus are frequently in segments ix, xi, or in xi only an unusual position. Those of the Australian 

 worms are often in segments ix, xii, also an unusual position. 



