34 E. R. HOSKINS AND M. M. HOSKINS 



size of the gland nor the histological structure in normal speci- 

 mens indicates that it is more active during metamorphosis than 

 before. However, when we consider that the thyroids do not 

 decrease appreciably in actual size during metamorphosis, where- 

 as the size of the entire animal decreases at this time from one- 

 half to two-thirds, the result is that relatively, the thyroid 

 becomes continually larger during this transformation of the 

 body, and the young frog has relatively much more thyroid 

 tissue than the larva. One precocious operated animal (fig. 20), 

 which metamorphosed earlier than the controls and at a smaller 

 size, had regenerated thyroids relatively larger than normal; and 

 Gudernatsch ('14) and others have shown that feeding thyroid 

 to tadpoles hastens their metamorphosis. In another experi- 

 mental animal, however, the thyroid that regenerated and per- 

 mitted metamorphosis consisted of a single gland with a volume 

 less than one-fourth the combined volume of the two thyroids 

 in the normal animal, and this regenerated gland did not have 

 a normal cellular structure nor very much colloid (fig, 70). In 

 the other experimental larvae in which the thyroids regenerated 

 the two glands varied in size considerably. 



These observations together with those of Gudernatsch indi- 

 cate that, although an abnormally increased amount of thyroid 

 hastens metamorphosis of the frog larva and although this trans- 

 formation will not occur if the thyroid is entirely removed (and 

 the larvae kept on a normal diet), still a thyroid gland much 

 smaller than normal was able to furnish the stimulus necessary 

 to bring about metamorphosis. This small thyroid may have 

 produced its secretion at a rate more rapid than normal, but 

 the need for added secretion ought to have caused an hypertrophy 

 of the gland. 



The part played in metamorphosis by the thyroid secretion 

 has already been discussed. From the result of the experiment 

 on forced metamorphosis in a thyroidless tadpole, it does not 

 seem that secretion is necessary for the atrophy of parts which 

 occurs during this process. 



Microscopic structure. The two thyroids of R. sylvatica are 

 well developed normally and contain colloid in larvae of 15 to 20 



