Studies of Tissue Growth 457 



of the sum of the squares of the errors of these two specific amounts 

 as given in the table. One finds on making the calculations that 

 the differences are scarcely equal to or even less than the probable 

 errors when comparing the groups regenerating one, two, four or 

 six arms. If any of these four groups be compared with those 

 individuals regenerating eight arms the latter group seems to 

 show a distinct advantage in its rate of growth which is almost 

 three times greater than the expected errors of the difi^erences in 

 the several cases. It may be stated, therefore, that the data for 

 Cassiopea shows no relationship between the degree of injury and 

 the rate of regeneration with the possible exception of those indi- 

 viduals regenerating all eight of their oral arms. The latter class 

 seems to grow new arms at a more rapid rate than do specimens 

 injured in any other manner. In consideration of the limited 

 number of individuals and the wide variability within the group 

 even this latter difference may not be generally found. 



Zeleny ('07) concluded from a study of a small number of these 

 medusae that the rate of regeneration was fastest in those specimens 

 having lost six of their oral arms and slower in individuals that had 

 lost more arms, as well as in those growing fewer. It is obvious 

 that my more extended series of experiments on Cassiopea fails to 

 show any such advantage for those regenerating six arms over 

 those growing eight arms. Since Zeleny's series is only one-third 

 as great as mine the differences in rates shown in his table are 

 probably, like mine, not significant. 



The lower half of Table VIII shows that after 24 days the dif- 

 ferences in rates between differently injured medusae are still not 

 significant. 



We may now proceed to a consideration of the regeneration 

 rates of the arms in brittle-stars injured to different extents. A 

 large number of these animals were lost in unsuccessful aquarium 

 experiments. Others were successfully kept in floating "live- 

 cars" where they flourished and increased in size. The "live- 

 car" experiments were started with 150 individuals, 135 of which 

 were available for final measurements. Zeleny ('03) experi- 

 mented on the brittle-star Ophioglypha and in fact on this form 

 first obtained the data from which he suggested the principle that 



