121 



very strong suspicion that the visual organs of vertebrates will not be 

 able much longer to hold the position of isolation so long conceeded 

 to them?" This suspicion once aroused, is fed and kept alive by a 

 slow train of accumulating evidence, that is gradually changing it from 

 a suspicion into a firm conviction. 



From the condition represented in Fig. 2, there is a decline of 

 the accessory vesicles; while the optic vesicles develop progressively 

 and become the paired eyes, the rest of the system degenerates. This 

 unexpected history is represented in Figs. 4, 5, 6 and 7. Within the 

 space of about two hours the accessory vesicles have entirely dis- 

 appeared. The process of disappearance begins behind and advances 

 forwards. 



Fig. 4, represents a stage which I estimate to be a half hour older 

 than that shown in Fig. 2. There are still six mesoblastic somites, 

 but the neural tube is closed further backwards and the optic vesicles 

 are better differentiated. The accessory vesicles, particularly in the 

 hindermost part of the series, are not so definitely marked oö, though, 

 it is still possible to see outlines of the original six. 



In Fig. 5, the system of accessory vesicles is present but shorter 

 there are now four vesicles present, the hinder two having disappeared 

 and, therefore, this is a more advanced condition, but, I do not think 

 there is more than a quarter of an hour difference, by the conventional 

 way of reckoning, between the two embryos, and it would be difficult 

 to say definitely which is the older. The optic system is more ad- 

 vanced in Fig. 5, but the walls of the neural canal are further ad- 

 vanced in Fig. 4. In both specimens, there are six mesoblastic so- 

 mites. These two embryos illustrate individual variation, and also 

 show that we cannot depend on the number of mesoblastic somites 

 to determine stages with any degree of closeness. The mesoblast is 

 not at this time the focus of developmental activity, and many changes 

 go on in the cerebral structures, while the number of mesoblastic so- 

 mites remains the same. Fig. 6 shows a slightly older stage in which 

 the optic vesicles are more bulging. The accessory vesicles are, as in 

 Fig. 5, four in number. 



In Fig. 7, almost the last vestiges of accessory vesicles are shown. 

 They are now reduced to two and the second pair are in process of 

 disappearing. This specimen has 8 mesoblastic somites and is esti- 

 mated to be in the 27 hour-stage of development. 



Figs. 2, 4, 5, 6 and 7 show a gradation of conditions in this 

 system as follows: — In Figs. 2 and 4 there are six accessory ves- 

 icles, and they extend over a space of six metameric divisions of the 



