448 SEMI-CENTENNIAL OF TORREY BOTANICAL CLUB 
However, the distinction between chance variation and dis- 
coverable differences according to position (differentiation) in 
such similar organs as flower heads in chicory is by no means clear. 
Partial, chance, or fluctuating variations are of such intergrading 
ranges that differentiation according to position would not be dis- 
covered by such indiscriminate and chance methods of study as 
have usually been employed. Considering the first terminals and 
the last laterals which bloom on a plant, there is, as a rule, a very 
marked difference. There is also a difference between terminals 
of various ranks. But for certain terminals the number of flowers 
per head is identical with that of certain laterals; one class grades 
into the other. A strict interpretation of differentiation, however, 
would perhaps lead one to the view that even the variation seen 
in any one day among closely homologous terminals or laterals 
produced on a plant is significant of still finer grades of differen- 
tiation. : ' 
Such a view would evidently recognize a symmetry of develop- 
ment giving primary, secondary, and tertiary branches. If such 
be the case, it is clear that the transition from one to the other is 
more continuous than discontinuous; at least the gradations are 
slight. Although differentiation is in evidence, and is readily 
discoverable, it is not obviously discontinuous. 
In chicory the repeated branching and deliquescent habit of 
growth with terminal and lateral branches of various ranks and 
series give opportunity for a full expression and development of 
marked differentiation of the sort discussed above. If the dif- 
ferentiation which does occur in chicory in development of number 
of flowers in such homologous organs as flower heads is any measure 
of differentiation between individuals, races, or species as such, 
it is indicative of continuity rather than discontinuity. 
In judging an individual as a whole for lower number per head, 
the collection of data at the beginning of the blooming season will 
give a different estimate from what would be obtained at the close 
of the blooming season. As a rule, higher values prevail in the 
first part of the blooming period. The variability observed will 
be less the shorter the period of time covered. There will be less 
variability with the limitation of data to terminals only, to laterals 
only, or to terminals or laterals opening during one day. 
