88 BULLETIN OF THE 



dorso-ventrally and gi'ows out to meet the invaginations of the ectoderm to 

 form tlie gill slits, there are left a few scattered cells (the upper portion 

 of those marked i. c. m. in Figure 5) between the gut and the neural 

 tube (compare i. c. m'., Figs. 6 and 7) ; but the most of these cells are 

 left to unite subsequently in one single mass below the gut and between 

 the edges of the two lateral plates (i. c. vi., Figs. 6 and 7). 



Thus the cells traced back in time from the endothelium of the heart 

 (Fig, 11 to Fig. 6) are the same as those marked i. c. m. in Figures 2 

 and 4. 



These cells are easily homologized with those described by Ziegler ('87), 

 Boyer (''J2), and others, and called by them the " intermediate cell-mass," 

 a term which I shall adoj^t in the further description. The problem be- 

 fore us is now reduced to the question of the origin of that part of the 

 intermediate cell-mass which is found in the region where the heart is 

 formed. 



In studying the origin of these cells I have made use of fundulus 

 as well as of the cod, the cells in the former being much larger and less 

 crowded together than in the latter, while the rate and manner of devel- 

 opment in both are veiy much alike. 



There are five possibilities as to the origin of these cells, and all have 

 been defended. The intermediate cell-mass may arise either from (1) the 

 parablast ; (2) the endoderm ; (3) the lateral plates; (4) the protover- 

 tebnx' ; or (5) from two or more of these. These possible sources will 

 be discussed in the order given. 



1. Parablast. — It seems as if the only circumstances under which 

 one could maintain a parablastic origin for the cells would be when 

 neglecting entirely the younger stages and arguing merely from the posi- 

 tion of the cells, as Hoffman ('82, p. 35) did. 



In such stages as Plate IV. Figs. 13-15 (fundulus), it is difficult to 

 conceive how the cells composing the intermediate cell-mass could have 

 passed up from the parablastic layer between the gut and the lateral 

 plates to the position they there occupy. In my opinion there is only 

 one possible explanation of this condition if these cells originate from the 

 parablast, namely, that before the formation of the posterior part of the 

 gut they migrate in from behind. But the study of series of longitudinal 

 and cross sections shows that the separation of these cells from the para- 

 blast is so complete, and for such a distance posterior to the region of tiie 

 heart, that to have all of the cells of the intermediate mass arise from a 

 proliferation of the parablast cells farther back would involve an inipos- 



