364 



OJf EAllTinVOllMS FROM BKITISH EAST AFKICA. [Apr. 16, 



earlier account of this species, and o£ a few corrections in matters 

 of detail. The larger of two examples examined measured 72 mm. 

 by a greatest diameter of 2 mm. The worm is thus quite as 

 slender as Gordiodrilus eUyans, and the more robust form of 

 the orighial specimens is perhaps merely a matter of greater 

 contraction. 



The openings of the two pairs of spermiducal glands is, as 

 correctly stated in the original description of this worm, upon 

 segments ^ xvii. and xviii. As to internal characters, it has been 

 noted that this is the only species of Gordiodri/us which possesses 

 a gizzard. The structure of this part of the alimentary canal 

 shows some further peculiarities which have not yet been referred 

 to. The gizzard in segment viii. has quite stout muscular walls, 

 but the lining of cuticle is not strongly developed as is the case 

 M'ith earthworms where the gizzard is a prominent structure. 

 Moreover, the gizzard by no means occupies the whole of the 

 viiith segment ; the last one-fourth or thereabouts is occupied by 

 a portion of oesophagus, which differs from other parts of that 

 tube in that the muscular layer is rather thick, about as thick as 

 the epithelium lining it. There is thus evidence that the gizzard 

 of this species is in a state of degeneration. In segment vii. there 

 is a similar thickening of the muscular walls of the oesophagus, 

 the layers being again about as thick as is the epithelium beneath 

 them. Here, therefore, is another, and a rudimentary, gizzard to be 

 taken account of. The species seems to be descended from some 

 form in vshich there were two gizzards in vii. and viii., and while 

 one of them has neai-ly disappeared the other is commencing to 

 undergo reduction. These facts further emphasize the bond of 

 union between the genus Gordiodrilus and its ally Nannodrilus, 

 though in a different way from the likenesses shown between 

 GordiodrihfS papillatus and Nannodrihis. The genus Nannodrilus 

 has tMo gizzards, which lie in the two consecutive segments vii. 

 and viii. The facts, however, must apparently be iiiterpreted on 

 the assumption that from Nannodrihis arose two separate lines of 

 descent, one represented by Gordiodrilus rohustus, from whicii 

 again G. dominicensis can be deri^■ed as well as perhaps G. diiheca. 

 The second line gave rise to G. papillutus in the first place, from 

 which may have arisen G. Unuis on the one hand and G. elef/ans 

 on the other. The relations of G. zanzibaricvs " are not so plain 

 as the others appear to me at present to be. 



^ It is necessary to emphasize this fact since some error has crept into my 

 original paper upon this genus in respect to the positions of the spermiducal 

 gland-pores. I find on re-examination of my preparations that in G. elegcn/s 

 the pores are correctly stated (upon pp. 84 and 9U) to be upon segments 

 xviii.. xix., and incorrectly stated (upon p. 95) to be upon segments XTii., 

 xviii. On the other hand, in G. doinlnicensin the same pores are, as in 

 G. rolnsi'Us, upon xvii., xviii., as correctly stated on the table on p. 95 of my 

 memoir; they are incorrectly stated upon pp. 91 & 94 to be upon xviii., xix." 



- As a small matter it may be well to note that " Gordiodrilus matthewsi,'' 

 spoken of on p. 453 of the Mon. Olig., is not, as Michaelsen has suggested, a 

 lapsus penn<e for G. rohuatus but for G. zanzibarieus. 



