402 On Alternate Generation of Annelids, 



of the different rings are so well developed that it can be 

 unmistakably recognized as a parent stock of Autolytus 

 with a small number of rings (see PI. X., figs. 13 and 14). 

 Slight swellings having made their appearance on the 

 sides of the first ring, then the second, and so on, and 

 being gradually changed into the dorsal cirri, as seen in 

 fig. 14, C, O', PI. X., the anal cirri increase in length ; and 

 by the time the last ring has a dorsal cirrus developed, the 

 middle tentacle (A) (Pi. XL, fig. 1) has again outgrown the 

 lateral tentacles, the tentacular cirrus ( O") has been devel- 

 oped, and the worm has all the appearance of a parent 

 stock with short cirri, ( C, C, C'^', O'). From this stage the 

 development goes on very rapidly, the number of rings 

 increases very fast, the tentacles and cirri grow to their 

 full length soon after, and in a couple of weeks the em- 

 bryo has passed through the stages represented in figures 

 2, 3, PI. XL, and has reached its full size, (PL X., fig. 1,) 

 when it is ready to begin the reproduction by division, as 

 I have described it above. 



If we compare the embryonic development of the parent 

 stock from the egg to the development of the different 

 parts of the male and female from the parent stock, we 

 cannot fail to be struck with their perfect coincidence. 

 The middle tentacle (A, a) is developed first in both, then 

 the lateral tentacles of the head (A', A'^, a', a'^). The 

 large eye is first visible, and then the smaller one ; it is 

 only when they are quite advanced that the long dorsal 

 cirri make their appearance in the embryo and in the 

 sexual individual ; the long setae of the latter developing 

 at a time which corresponds to the formation of the stout 

 bristles in the embryo ; thus showing a perfect correspond- 

 ence in the successive development of the different parts in 

 these two modes of reproduction, and giving us a true 

 perception of the value of embryological characters for 

 classification. We cannot fail to see that the stages 



