558 S. Morgiilis 



tail) was not utilized in these experiments. Although the average 

 length of these terminal pieces was somewhat over forty to forty- 

 five segments they all died. 



Pieces of corresponding levels were kept in the same dish, 

 so that they were all practically under like conditions. Accord- 

 ing to the level to which the pieces belong, they will be named 

 Ai, Aj, A3, A4, A5, Ag. At the end of two weeks all the 

 tails that had been regenerated by these pieces, were cut off; the 

 number of their segments as well as the number of segments in 

 pieces, by which they were produced, was recorded, and the results 

 are given in Table III. In this table is also given the average num- 

 ber of regenerated segments per one old segment (in the last line). 

 By this method of calculation all the individual variations were 

 obliterated, and at the same time this, so to speak, "ideal old seg- 

 ment" with the corresponding "ideal" number of new segments 

 served to indicate the regenerative power at a given level, and 

 served also as a basis for the comparison of the rates ot regenera- 

 tion. 



In accordance with our method of calculation, we find that to 

 each of the original segments at the first level, there were on the 

 average 3.2 new segments; at the second level 3.3, at the third level 

 3.1, at the fourth level 2.6, at the fifth level 1.8, and at the sixth 

 level only 0.9 new segments. 



There is no evidence of a correspondence between the number 

 of old segments and the number of new, regenerated segments, 

 provided the old parts are not very different in length. Glancing 

 over theTable III we can see at once that in pieces at the first level, 

 those having twelve segments produced thirty-two or fifty new 

 ones; those having thirteen old segments thirteen or forty-two. 

 In pieces at the second level also, those of eleven old segments 

 regenerated twenty-nine or fifty-six, and those of sixteen may 

 regenerate eighteen or fifty-four new segments, etc. The same 

 lack of correspondence will be found in all parts of this table as 

 well as in subsequent ones. 



On the other hand if we compare the smallest pieces of a cer- 

 tain level with the largest pieces of the same level we may some- 

 times find very considerable differences in the amount of regener- 



