REGENERATION OF FINS OF FUNDULUS HETEROCLITUS. 39 1 



A second experiment was started on August 2 and continued 

 until September 5 a period of four weeks and five days practic- 

 ally the same as those in Table I. In this case the fish were fed 

 regularly for it was desired to test the effect of difference in food 

 on regeneration. In Series D the caudal alone was cut, in E the 

 caudal and right pectoral, in F the caudal and left pectoral, 

 lowing is the table of results, Table II. 



We may arrange the results as before and have the following 



table. 



Since but one pectoral fin was cut off in both E and F it is 

 clear that we cannot compare difference in regeneration in those 

 two lots since the degree of injury was the same. In this case 

 we compare E with D and F with D. Here is the first excep- 

 tion to the uniform results obtained in series A, B and C from 

 Table I. For on comparing E with D we find that the mean 

 difference in regeneration is more than three times the probable 

 difference, hence according to our formula regeneration is really 

 greater in D than in E. At first we might think that since the 

 Two differ as to degree of injury that this must be the cause of 

 the result that is, that in this case at least the least injured 

 regenerated more in the same time. But is this the case ? For 

 comparing D and F (the injury to F being the same as that to 

 E) we see that the mean difference between the two is less than 

 the probable difference which shows that no importance can be 

 assigned to that mean difference and that D and F regenerated 

 practically the same amount in the same time. The case of D 

 and F confirms the result already found in A-B, B-C and 

 A-C as to caudal, and in case of B-C as to pectoral. The case 

 of D-F also shows that whatever is the cause of the discrepancy 

 between that and D-E we cannot say that E regenerated less 

 because it was injured more than D. 



