444 Max Mapes Ellis 
was proportional to the amount removed. In Experiments 
2 and 3, the same relation existed between the amount regenerated 
and that removed, at the end of the twelfth day as was noted in the 
two previous experiments; this relation, however, did not obtain at 
the end of the fifth day after the operation. Subsequent experi- 
ments gave two important facts in this connection. (1) Regenera- 
tion ceases before the part removed has been completely regener- 
ated; and (2) the amount regenerated is proportional to the amount 
removed, regardless of the absolute amount removed, only at the 
time regeneration ceases. [he tadpoles from which these data 
were collected did not grow while under observation, as shown by 
the control series. In but a single case did the tadpoles used 
regenerate the complete amount removed; these animals were but 
a week old and growing rapidly (Experiment 10). ‘There is a 
very probable explanation for this incomplete regeneration on the 
basis of the state of development of the tadpole. Kammerer 
found that the rate of regeneration of tadpole tails varied not with 
the absolute age of the tadpole, but with its state of development. 
Suppose then, that there is a time (as there probably is) while 
the tadpole is very young, that it is capable of regenerating com- 
pletely a part of the tail lost by injury. Data collected from obser- 
vations on various animals show younger animals, as a rule, to 
regenerate more readily than older ones; that is, animals at a 
more advanced stage of development regenerate less rapidly. This 
fact was pointed out as early as 1786, by Broussonet in his obser- 
vations on the regeneration of the tails of fishes. It is also known, 
that the adult frog, normally, does not regenerate. ‘The tadpole, 
then, must lose its ability to regenerate in one of two ways: either 
suddenly, at some particular point in its life; or gradually through- 
out its larval life. The latter is the more probable, and is sub- 
stantiated by the data collected. If the ability to regenerate be 
dependent upon the state of development it were possible then to 
establish a “regenerative coefficient”’ for the various ages of tad- 
poles; which coefficient would be the maximum per cent that 
the particular stage is capable of regenerating, under the best con- 
ditions for regeneration. ‘This per cent, naturally, however, 
would not be regenerated; for various other factors also control the 
