KI^IGHA MITCHELI. SCIENTIFIC SOCIETY. 71 



Hertwio- in his text book, there very early developes a 

 sickle-shaped thickening' which lies between the area 

 pellucida and the area opaca, in the po sterior reoflon of 

 the blastoderm. A g-roove, the sickle g'roove, is present 

 in this thickening, and in the median line there is a short 

 anterior projection called the sichel-kuopf. The prim- 

 itive streak is produced b}' the continuous g-rowth in 

 the median line, of the sichcl-kuopf\ and is therefore 

 an outg'rowth of the sickle. Since neither the sickle 

 nor the primitive streak is at an}- time connected with 

 the blastoderm edg-e, the latter structure cannot be 

 reg-arded as a part of the blastopore, which is repre- 

 sented exclusiveU^ by the two former structures. 



The, contradiction between Duval's and Roller's ac- 

 count concerns a fundamental feature of the process of 

 g-astrulation, and more facts on the early history of the 

 bird blastoderm are much to be desired. Duval him- 

 self, in his criticism of Roller's papers (1. g.), states it 

 as his opinion that the sickle is an inconstant feature, 

 of no morpholog-ical importance, belong-ing- in the same 

 categ'ory as other local thickening-s of the blastoderm. 

 I may mention that I have myself looked througfh very 

 3^oung- blastoderms, in which the primitive streak was 

 from one-half to two-thirds the leng-th of the area 

 pellucida, without discovering^ in the majority of them 

 any trace of the sickle. I am aware that Roller 

 describes the sickle as becoming' much less conspicuous 

 with the continued g'rowth of the streak, but his fig-ures 

 of blastoderms^ corresponding- in ag'e to mine, show an 

 evident remnant of the sickle, while I can find no trace 

 of such a structure in the majority of ni}' embryos. 



Roller, it will be remembered, kept his eg'gfs at a 



4. SB. d. Konig-. Akad. d. Wiss. 1879, Beitrag". &c.. fig-s. IV. a. IV. b. V 



