390 Miscellaneous. 



Krohn no doubt committed the same error. The larvae are certainly 

 on the ventral side, as (Ersted saw in his E. naidina. 



I have frequently met with E. gemmifera, male and female, in a 

 state of sexual maturity. No error was possible ; and in both I have 

 found individuals with long setae and others which were destitute of 

 them. I know very well that it is said these setee may become de- 

 tached in the natatory movements of the animal, and I have seen 

 specimens in which this had probably taken place. But when all 

 the segments are absolutely destitute of setse doubt is no longer 

 possible, and the more because we observe no trace of the implanta- 

 tion-sacs of the capillary seta. Thus the principal reason which 

 guided M. Pagenstecher disappears. The following is the course of 

 development : — We see an ovum originate at the posterior surface of 

 each dissepiment, starting from the tenth segment (the last three or 

 four segments remaining free), on each side of the median line and 

 below the intestine. The two ova remain always alone in the seg- 

 ment, as indicated by Claparede in his Pcedophyliuv , which is perhaps 

 identical ; they enlarge until they press the intestine upwards and 

 meet in the median line. Sometimes one of them passes in front of 

 the other, and, their envelopes being very flexible, they mould them- 

 selves upon the ani'ractuosities of the cavity of the segment. After 

 deposition the ovum is attached by a very distinct peduncle to the 

 base of the ventral cirrus. The line of separation is easily seen when 

 the animal is observed from the side and without compression. The 

 ovum is deposited before any segmentation has taken place. The 

 segmentation appears to be very regular ; and all the ova pass through 

 the same phase nearly at the same time. As these little annelids 

 die very quickly in captivity, a great number of individuals are neces- 

 sary in order to observe the successive phases. In the last stage 

 observed in the egg the segmentation was complete, and the ovoid 

 larvae showed the buccal orifice distinctly. They exactly filled the 

 envelope ; and it is no doubt their increase in size that causes its 

 rupture. The larvce are already naked at the moment when they 

 present the form of an ovoid mass with clear ectodermic cells and 

 strongly coloured endodermic spherules, without the least trace of 

 transverse division. The endodermic mass is cordately emarginate 

 on the side towards the mouth. 



These larvee, which are very convex on the back, show at their 

 free extremity three small scarcely perceptible ectodermic buds, 

 representing the first traces of the tentacles ; two other exactly 

 similar buds, situated at the other extremity, will become the anal 

 cirri, which, in the course of their development, pass over on each 

 side of the ventral cirrus of the mother. The point of li: lion of 

 the larva is therefore exactly that of the egg. When it becomes 

 detached we do not see the peduncle described by ragenstecher, but 

 a slight impression at the level of the anus, which perhaps acts as a 

 sucking-disk. The young larvae enlarge regularly, and do not begin 

 to bear setae until there are already four or five segments between 

 the head and the anal segment. * Thus, just as in the Autohjtus 

 investigated by A. Agassiz, there is no development of cinctures of 



