DEATH AND THE DURATION OF LIFE 93 



however, a long life and continue to grow beyond the 

 usual size of the tadpole. When, however, such super- 

 annuated tadpoles are fed with thyroid they promptly 

 undergo metamorphosis. These observations cleared up 

 an old biological puzzle. Salamanders also go through a 

 metamorphosis which is, however, less striking than that 

 of the tadpole of a frog. In the salamander the metamor- 

 phosis consists chiefly in the throwing off of the gills, and 

 in changes in skin and tail. In Mexico a salamander oc- 

 curs which through its whole life maintains its tadpole 

 form, namely the axolotl. Attempts to induce the axolotl 

 to metamorphose failed until after Gudernatsch's discov- 

 ery an investigator fed the axolotl thyroid gland, and this 

 brought about metamorphosis. The thyroid gland stores 

 the traces of iodine taken up in our food and it seemed 

 possible that the iodine contained in the thyroid was the 

 active principle causing metamorphosis in tadpoles. This 

 was confirmed by Swingle who succeeded in inducing 

 metamorphosis in tadpoles by feeding them with traces of 

 inorganic iodine. According to our present knowledge, 

 the duration of the tadpole stage seems to be the time re- 

 quired to store the necessary amount of certain com- 

 pounds, one of which contains iodine. 



Insects, like the fruit fly, hatch from the egg as mag- 

 gots which grow at the expense of the food they take up 

 and which, at a certain age, metamorphose into a chrysa- 

 lid ; and from this- chrysalid at a given time will rise the 

 winged fly. Feeding of thyroid to the maggots of the 

 fruit fly will not accelerate their metamorphosis, and we 

 can not tell whether in this case metamorphosis is due to 

 the accumulation or formation of a definite compound in 

 the body, though this may well be the case. The idea 

 presented itself whether the duration of the larval or 

 maggot stage was not also determined by the temperature 

 (provided the food supply was adequate). We meas- 

 ured, therefore, the influence of temperature upon the 

 duration of the larval state in aseptic fruit flies i.e., from 



