MUSEUM OF COMPARATIVE ZOOLOGY. 91 



lateral plates, although for regions anterior to this he gives a different 

 origin, and finds no evidences of a forward migration of the cells arising 

 from the lateral plates. I have looked carefully at the positions of cells 

 and for other indications of such migration, but have failed to find any 

 and am convinced, whatever may he the condition in a region posterior 

 to the heart, that it is impossible — owing to the distance between the 

 two regions and the lack of evidence of migration, together with the 

 great number of cells in the intermediate cell-mass of tlie heart region — 

 for the entire intermediate cell-mass, and particularly that of the heart 

 region, to have arisen from tlic lateral plates. 



4. Frotouertehrce. — However unconvincing the arguments against 

 these first three possibilities of origin may be, I feel that the evidence in 

 favor of the fourth possibility — origin from the protovertebrse, or, in 

 earlier stages, the protovertebral plate — is strong enough to prove con- 

 clusively that the intermediate cell-mass in the fishes which I have 

 studied takes its origin in the heart region from that source. 



Unlike the parablast the protovertebra3 are from the beginning in di- 

 rect continuity with the intermediate coil-mass, except in stages of about 

 the fifth and sixth day in the cod (such as are shown in Figs. G to 8), 

 and even then it is to be seen that the chief portion of the intermediate 

 cell-mass is connected witli the protovertebrse by a slender cord of cells. 



Unlike the endoderm and lateral plates, the protovertebra3 are at all 

 stages poorly defined on the side toward the intermediate cell-mass. 



This is especially the case in such stages, for instance, as the one 

 sliiiwn in Figure 2. It is evident from the mutual relation of the parts, 

 tliat at this stage the lateral plates could not be supposed to have pro- 

 duced the cells which extend along the lower surface of the neural tulie 

 as far as the chorda. Only two cells in this figure (/. c. mJ) suggest by 

 their position the possiliility of an origin from the lateral ])late, whereas 

 the main portion of the mass, which 1 have at this early stage desigrated 

 as i. c. m., is clearly continuous with the lower margin of the protoverte- 

 brje, from which, indeed, they are not yet wholly detached; moreover, 

 the two cells mentioned are more likely to have originated from the 

 protovertebrse than from the lateral plates, because tlie latter are now, 

 and continue to be, more sharply defined than the lower margin of the 

 protovertebrai. It must be admitted that at a later stage, owing to the 

 folding in of the endoderm to form the fore-gut, the median margin of 

 the lateral plate is brought into a position more compatiljle with its fur- 

 nishing cellular elements to that part of the intermediate coH-mass which 

 lies below the head-gut, and from which the organs in question arise. 



