24 E. R. HOSKINS AND M. M. HOSKINS 



histological differentiation, and appears to be the most actively 

 secretory portion of the entire hypophysis. Atwell ('18) names 

 the lobes of the frog's hypophysis pars tuberalis, pars intermedia, 

 and anterior lobe proper. These names correspond to those 

 used in higher groups of animals, but we have used the terms 

 anterior, superior, and inferior as being more descriptive. The 

 anterior lobe is very inconstant in size and shape. In some 

 cases it is forked, and it may be completely subdivided into two 

 smaller lobes in later larval stages and frogs. Its most usual 

 arrangement is that of a thin plate of cells closely adherent to 

 the infundibulum, anterior to the inferior lobe. At times it is 

 completely buried within the inferior lobe. It showed no par- 

 ticular secretory activity in our microsopical preparations and 

 may not be of much importance, although its size was usually 

 proportional to that of the body of the animal. After studying 

 and drawing with a camera lucida the hypophyses of more than 

 ninety larvae and young frogs and after careful comparison of 

 the drawings with each other, we selected figures 36 to 45 as 

 properly representative of the hypophyses of the different groups 

 of animals. The glands were studied microscopically in other 

 specimens also. The hypophysis is too irregular in outline to 

 permit accurate determination of its volurrfe from measurements 

 of the three principal diameters, as was attempted by Rogers 

 ('18), and is too small for accurate weighing or direct volume 

 determinations. 



It is obvious that glands vary in size with the size of the body 

 in animals in the same stage of development. Rogers' three 

 groups of animals were not of the same size nor of the same shape, 

 and since he failed to reduce his measurements to comparable 

 figures, his conclusions are unwarranted and, as we shall show, 

 are mostly incorrect. His experimental larvae, judging from 

 differences in R. sylvatica with which we have worked, must 

 average nearly, if not quite, twice as large as his control larvae, 

 but their body length is only about 25 per cent greater (Rogers, 

 p. 594). Dividing the volume of the glands by the body length 

 makes the glands of the longer animal appear relatively large. 



