MUSEUM OF COMPARATIVE ZOOLOGY. 79 



the embryo becomes more folded upon itself ventrally, as shown in 

 PI. II. fig. 10, and the legs, increasing in length, gradually approach and 

 finally overlap each other in the median line. The embryo has no\y 

 acquired a strong ventral flexure — the reversion is completed. 



During this period the bases of the cheliceras in growing have moved 

 forwards and met in the median plane, so that they appear as pre-oral 

 appendages. There has also appeared between their bases a prominent 

 outgrowth to form the rostrum. 



Balfour ('80, p. 180) endeavors to account for the process of rever- 

 sion as the result of a rapid " elongation of the dorsal region, that is, the 

 region on the dorsal surface between the anal and the procephalic lobes." 

 I understand by this that it is to the growth of the ectodermic cells of 

 the dorsal region that he would ascribe the elongation of the dorsal surface. 

 I shall endeavor to show presently that this explanation is not sufficient 

 to account for the changes which actually take place during reversion. 



The growth of the derivatives from the ectoderm during the period 

 under consideration is very great. At the beginning of the period the 

 stomodseum forms a pocket-shaped invagination with a small external 

 opening. Its calibre diminishes, except at its anterior end ; it continues 

 to grow inwardly, and at length forms an arched tubular organ, with its 

 free end directed backward, and projecting some distance into the yolk. 

 Near the close of the period its deep end becomes somewhat enlarged to 

 form the rudiment of the sucking stomach. To the latter are attached a 

 vertical muscle (mu. vrt. PI. IX. fig. 62) extending to the dorsal wall of 

 the embryo, and two lateral muscles (mu. lat.). 



The proctodajum is a later formation, which makes its appearance as an 

 infolding at the tip of the tail-lobe some time after the beginning of thi's 

 period. The relation of the tail-lobe to the rest of the body is best ap- 

 preciated from sections, since it is not always evident from surface views 

 that there is a deep fold which serves to separate it from the underlying 

 portion of the dorsal surface. The prominence which it attains and the 

 changes which it undergoes are readily traceable in a series of figures 

 from successive stages during reversion (PI. VIII. figs. 50-54). The 

 strong resemblance of the condition shown in Fig. 50 to that which 

 Bobretzky ('74, Fig. 15) has figured for Oniscus at an apparently similar 

 stage of development, misled me into the supposition that I should find 

 the proctodaeum of Agelena developing in the manner described by him 

 for Oniscus. But such is not the case. In Agelena the tip of the lobe 

 is the tip of the tail — the morphological end of tlie body, and the 

 depression which separates this lobe from the neighboring portion of the 



