291 



Curves obtained by this method are shown in figure 6. If we 

 compare the line of regeneration after the cut at level y with that 

 after the cut at level ß, we find that, although the curves terminate 

 at the same points, they pursue strikingly different courses, so that 

 whereas after the ß cut only 4 or fewer rings were regenerated in 

 only 12 °/ of the cases; after the y cut the same number of rings 





















SI 



// 



/ 

 / 



















/ 

 / 



















/ /' 



S 















.•**' 







y© 











,15 ■*»■' 



2«, 















' 



ap 





dy 











// 



/ 



















/ 



/ 

 1 

 i 

 /' 









6.6, 









y 







& } 



>y 



if 







,,-'' 



-' 







/ 



/ 





r ,fS . 



i s 



.''' 















aj! 



r ctJ 



.s^JZ 













ly'' 



*' 













Regen, o 

 Regen ol 



r rings m 

 inpsnot 



'fAro 

 tnse$. In 



"cuts. 

 et,a/kuls. 

















1 





10% 10% 50% 'tO% SOH ec% 70% SOh so* m% 



Fig. 6. Curves of Distribution of Regeneration 



regenerated in 74°/ of the cases. These curves also show, on the 

 other hand, the lack of definiteness of the result after a cut a any 

 level. 



5. As has just been observed, the curves (fig. 2) are not wholly 

 regular. The principal maxima occur at 4 riDgs (y cut), 6 rings (a, 

 ß cuts), 9—10 rings (a, ß, y cuts). Uniting all cases of regeneration 

 (622 cases, a number large enough to smooth out insignificant irre- 

 gularities) we still find two prominent maxima in the curve; namely, 

 one at 4 and one at 9 rings (fig. 5). The significance of these two 

 maxima becomes clear when we draw separately the curves of cases 

 in which rings only were produced and in which an uns eg- 



