ORIGIN OF THE LATERAL LINE PRIMORDIA 613 



broad and is again continuous with the more ventrally located 

 angular thickening. Each contact of the endoderm with the ec- 

 toderm is placed diagonally on the body with the anterior end of 

 the contact lower than the posterior end after the second con- 

 tact is formed. The first becomes free at the posterior end, 

 where it extends back of the area of contact. 



This process is repeated in the case of the remaining gills. 

 Each gill thickening is continuous at first at its posterior end with 

 the angular thickening, later becomes separated near its anterior 

 end from the more ventral angular thickening and finally also 

 at its posteror end, after which it remains some time as a dis- 

 tinct thickening, at least until after the formation of the epi- 

 branchial ganglion of each gill, and then later disappears. 



The gill thickenings are placed diagonally in the body with the 

 anterior end lower than the posterior, so that their general re- 

 lation to each other is like that of the shingles on a roof. This 

 relation makes clear the splitting of each thickening at its anterior 

 end before the thickenings of two successive gills become detached 

 at their posterior ends. 



The gill thickenings are to be looked upon as the ectodermic 

 portions of the gill which is formed before the contact between 

 ectoderm and endoderm is made, just as the endodermic gill 

 pocket is in process of formation before the contact stage is 

 reached. The import of the angular thickening is less easy to 

 understand. It may have no significance further than its rela- 

 tion to the gill thickenings, or it may have something to do with 

 strengthening the body at the point where the vertical lateral 

 body wall turns and becomes horizontal as it spreads over the 

 yolk. It seems not to be present after the yolk is further 

 absorbed and the gills have formed, and its appearance is 

 closely associated with the appearance of the gill slits. The 

 position and relation of both the angular and gill thickening to 

 surrounding structures are such that unless they are followed 

 closely they might be mistaken for the postauditory placode or 

 the primordia of lateral lines. 



5. The posterior extensions of the gill thickenings, as well as 

 the epibranchial placode, are closely associated with gill thicken- 



