MAMMALIAN CESTODES. 557 



consideration of the disposition of the eggs in proglottids anterior 

 to the two that have just been described bears out this view. 

 In the two j^roglottids anterior to the two terminal ones, the 

 arrangement of the eggs, or rather developing embryos, in the 

 medullary parenchyma was as follows : — The medullary region is 

 not so entirely given up to these eggs as it is in the case of the 

 last two segments that have been described above. In the two 

 latter but slight traces are left of the testes ; in the less matuie 

 proglottids moi'e testes are found. They have, however, com- 

 menced to disappear, for in the proglottids antecedent to these the 

 testes are more abundant and therefore take up more room. The 

 space apparently thus left by the degeneration of the testes is 

 occupied by the developing eggs. It is not, however, a continuous 

 cavit}' running from end to end of the proglottid (see text-fig. 89). 

 It is divided here and there by delicate trabeculse into compart- 

 ments, each of Avhich contains a considerable and varying numbei" 

 of developing embryos*. Further forward still, the embryo-con- 

 taining spaces are less conspicuous owing to the full development 

 of the testes, which thus occupy more room. I can draw no line 

 of demarcation between these cavities in which lie the developing 

 embryos on the one hand, and spaces which lodge ova and form 

 actually a part of the ovary on the other hand, in the more 

 mature proglottids which have begun to become elongated. 

 As already mentioned in describing the ovary of this species 

 of Dijjylidiuvij the ovaries are mature or, at any rate, 

 possess plenty of full-sized ova in proglottids which are not 

 at all elongated and are, indeed, still broader than long. In 

 these proglottids, however, there is no trace that I could find 

 of a definite uterus, and nothing at all like the obvious retiform 

 uterus of not fully mature proglottids of D. ccmhium. However 

 we maj interpret the embryo-holding spaces of D. dongolense, it 

 is clear that they cannot arise in the same way that the egg- 

 sacs of D. caninum arise, i.e., by a breaking up of a continuous 

 I'etiform uterus into a number of discontinuous sacs. 



With what, then, can we compare the spaces which surround 

 the developing ova in this and other species of Di2')yUdmm — in 

 fact, possibly in all with the exception of D. caninum ? I do 

 not think that any genus has been described as possessing an 

 exactly similar mode of lodgment of the developing eggs. But 

 there are nevertheless some comparisons to be made with other 

 forms. 



The final stage shown in Dipyliditom dongolense, where the 

 eggs ai-e uniformly scattered through the medullary region, each in 

 its own separate compartment, is suggestive of Ooc/io7-istica, with 

 which genus Dipylidium is usually considered to be properly 

 associated. The comparison, hoAvever, can hardly be exact, since, 

 as we have seen, this final stage is preceded by larger spaces con- 

 taining each of them many developing embryos. It seems to- 



* In one specimen I have also seen ova enclosed bj' threes and fours in cavities in 

 the cortical layer. But this seems exceptional. 



