No. 3-] THE VERTEBRATE HEAD. 557 



The optic vesicles are formed first, and when, at a very little 

 later stage, the others arise behind them, it appears as if the 

 process of eye-formation were repeating itself serially. 



These circular pockets not only arise in a similar way, but 

 structurally resemble the optic vesicles. In cross and longitu- 

 dinal sections the cells of the sunken patches are similar to 

 those in the eye pockets. They may be designated accessory 

 optic vesicles. 



The assumption that they are primitively visual in character 

 may be going too far, but the basis upon which it rests will be 

 shown directly. The least that can be said of them (from their 

 structure, the place and manner of their appearance), I think, 

 is that they are segmental sensory patches. 



These structures begin to appear soon after the eye vesicles. 

 There are several pairs of them stretching in rows back of the 

 eyes, as stated above, the two anterior pair are the most prom- 

 inent. They gradually grow fainter ; the anterior one is the 

 best differentiated ; the second one is not so clear, and the 

 succeeding ones grow fainter as we pass backwards. I have 

 counted as many as four pairs in surface study ; but sections 

 show that the series is more extensive and contains not less 

 than eight pairs of these circular pits. The two anterior ones 

 are well shown in Figs. 9 and 10. These are the earliest surface 

 indications. In Fig. 17, which is slightly more advanced, the 

 accessory structures are clearly developed ; the figure shows 

 four pairs behind the eyes. In longitudinal sections of the 

 same specimen, PI. XXIX, Figs. ^6-%^, eight pairs may be 

 counted. The form of the depression, as seen at first from 

 the surface view, is ovate ; they soon become circular patches, 

 and form themselves into concave, shallow cups. I have seen 

 them in many specimens. 



It will be noted that they make their first appearance while 

 the neural plate is broadly expanded, and spring into promi- 

 nence while the neural folds are growing upwards. It is during 

 the rise of the neural folds that similar, but fainter, vesicles 

 can be made out, in surface study, behind the two anterior pairs. 



PI. XXX, Figs. 88 to 112 show twenty-five transverse sec- 

 tions of an embryo very slightly older than that represented in 



