192 Observation of the Origin of Varieties. 
effected nothing more than the isolation of the individ- 
uals of the new race. 
We have therefore to examine the "late summer" 
curves of the 13-rayed, 21 -rayed and double races. Let 
us begin with the first. 
The extreme limits of the curve of this race at the 
beginning of August were 11 and 21 ray-florets. These 
numbers gradually decreased until September, when only 
heads with 13 and 14 rays were formed. In the next 
year at the end of July the limits were 10-19, but in 
August 12-14. 
I examined the 21 -rayed race, with reference to this 
character, in the summer of 1898, dealing with the indi- 
viduals which had been saved for seed. The data for 
three of the plants 1 are summarized in the following 
table : 
PLANT FLOWERS NUMBER MIN. MED. MAX. 
A. Terminal 1 48 
September 1st 32 29 33 45 
October 10th 42 18 27 36 
November 1st 28 19 26 31 
B. Terminal 1 35 
September 1st 36 24 28 36 
October 10th 33 16 22 27 
November 1st 23 15 21 25 
C. Terminal 1 46 
September 1st 14 26 28 35 
October 10th 18 18 26 30 
November 1st 8 21 23 28 
We see that the numbers gradually shift in the direc- 
tion of the maximum at 21 (in the case of one plant 
actually reaching it), without any indication of the max- 
imum at 13 of the other race. The plants dealt with, 
therefore, clearly belonged to the 21 -rayed race. 
1 Over het periodisch optreden dcr anomallen op monstreuze 
planten. Kruidkundig Jaarboek, Gent, XI, 1899, pp. 57-58. 
