1902.] NATURAL SCIENCES OF PHILADELPHIA. 89 



made from a series of sections of this stage. These two divisions 

 of the anterior end of the gland are separated by a vertical parti- 

 tion of mesoblast. The two parts of the thyroid are more widely 

 separated by the mesoblast than they will be at later stages, but at 

 this stage the separation extends only a short distance from the 

 anterior end of the gland. The thyroid is here not so closely 

 pressed against the pharynx as was the case in previous sections, 

 being separated from the wall of the pharynx by a collection of 

 mesoblast cells. The cavity of each half of the gland is crescentic 

 in cross section, and on the left side is seen the beginning of the 

 lateral invagination. 



The thyroid cavity is closed in, now, for about two-fifths of its 

 entire length, both anteriorly and posteriorly, so that its opening 

 into the pharynx is reduced to a slit extending only one-fifth the 

 length of the gland. 



Sections through the thyroid, anterior to this opening (fig. 2b) 

 but posterior to the divided portion of the anterior end, show the 

 ca\aty to be considerably complicated by the thickening and folding 

 of the Avails. The simple, circular cavity of the preceding stage 

 (fig. lb) is now reduced, by the growth and invagination of the 

 Avails, to a narroAV vertical slit Avith lateral diverticula aboA'e and 

 beloAv, giving the cavity, in cross section, somewhat the shape of a 

 distorted letter T. The dorsal Avall of this caAdty (tig. 26, d.L) is 

 comparatively thin, being composed of a single layer of rather 

 short columnar cells, and is Avhat Dohrn calls the ' ' Decklamelle. ' ' 

 The ventral Avail is also comparatiA'ely thin, and shoAAS in the 

 centre, at the extreme anterior end of the unpaired cavity, a slight 

 upward projection (fig. 26, m.L), Avhich Avould seem to be the 

 first indication of the medial longitudinal partition that will later 

 divide the gland into two lateral parts. According to Dohrn, this 

 medial lamella is formed as far back as the opening of the gland, 

 before the lateral iuA'aginatiou takes place, but in my material this 

 is certainly not the case, as is s ho ami in fig. 2b. 



The lateral invaginations extend posteriorly to a point a little 

 behind the anterior edge of the slit-like opening into the pharynx, 

 fig. 2c being from a section that cut through the extreme posterior 

 part of the iuA'agination in the right side, but Avas too far posterior 

 to cut that of the left side, either because the section was somewhat 

 oblique or because the right invagination had proceeded further 

 than'tlie left. 



