REGENERATION OF FINS OF FUNDULUS HETEROCLITUS. 393 



partment in which they were kept during the five weeks they 

 were under observation. If the results in case of G H and I 

 confirm the results in cases A B, A C, B C, D F, etc., then I 

 think we are warranted in concluding that some other condition 

 other than degree of injury was responsible for that discrepancy 

 in case of E. The fishes in this third experiment were fed every 

 day or so. Following is Table III. giving the measurements of 

 these fishes. 



Treating these results as before we have the following table : 



We thus see that in this case also the mean difference is not 

 great enough to ascribe any importance to it. In each case it is 

 less than the probable difference and hence although the series 

 differ as to degree of injury, they do not differ as to rate of 

 regeneration. We can also tabulate the comparative regenera- 

 tion in the case of the pectorals I and H. 



In this case also the mean difference is less than three times 

 the probable difference which but corroborates the above con- 

 tention. We therefore find that in nine cases the result is that 

 although the fishes differed as to degree of injury, the rate of 

 regeneration was the same. The experiments are certainly varied 

 enough to represent different lengths of time, etc. There is but 

 one exception to the results and that we have shown cannot be 

 due to difference in degree of injury. 



