MUSEUM OF COMPARATIVE ZOOLOGY. 81 



The spinning glands are not yet definitely established, but in the anal 

 region on the ventral side of the proctod?eum there is a large accumula- 

 tion of ectodermic cells (PI. XI. fig. 70) from which they are subse- 

 quently developed. 



Late in this period the infoldings for the lungs arise. There appear a 

 pair of large oval masses of cells, the nuclei of which are arranged in par- 

 allel lines (PI. XI. fig. 73). From these cells the respiratory lamelLne of 

 the lungs are finally formed. 



The mesoderm likewise has been growing rapidly during reversion. 

 In the previous period it was confined to the ventral portion of the em- 

 bryo, but during the present period it grows upward on either side until 

 it reaches the dorsal median line, thus forming a continuous layer be- 

 neath the ectoderm, as well as an investment for all organs, which -arise 

 as outgrowths of either ectoderm or entoderm. 



The dorsal growth of the rudimentary terga, already spoken of as 

 external features, is followed by this underlying layer of mesoderm. 

 Early in the formation of the dorsal elements tliis niesodermic layer is 

 divided into corresponding somites. Balfour ('80, p. 181) concluded 

 that the cells out of which are formed the dorsal somites of the meso- 

 derm " are not derived from prolongations of the somatic and splanchnic 

 layers of the already formed [ventral] somites, but are new formations 

 derived from the yolk." My sections, however, indicate that there is a 

 direct continuity between the two (PI. IX. figs. 59, 61), and that the 

 dorsal mesoderm is an outgrowth from the previously established ventral 

 niesodermic somites. 



It is during this period also that the heart is formed. "While I have 

 been unable to arrive at an entirely satisfactory understanding of the 

 details of its formation, I am convinced that it is not, as Balfour states, 

 developed from a solid cord of cells, but from the dorsal limb of the up- 

 growing mesoderm, and tliat its dorsal wall is closed first, while the 

 ventral wall — the floor — remains for a time widely o])en below, thus 

 communicating freely with the yolk. My sections also sliow that at a 

 later period the aorta is formed, by means of a Ci^nstriction, from the 

 mesenteron. This agrees with the recent observations of Schimkewitsch, 

 ('84"). 



A layer of characteristic cells, to which Balfour alludes in speaking of 

 the formation uf the dorsal mesoblast, precedes the formation of the 

 heart in tlic dorsal region. These are what have been called " primary 

 entoderm " cells, and are sharply distinguished from the surrounding cells 

 by their large size, their large, oval nuclei, and their yellowish tint. Tliese 



VOL. XII. — NO. 3. G 



