chromosome morphology. This is certainly one of the 
most important types of evidence in a problem of this 
nature. However, it is not the only evidence available 
and it may be hazardous to overlook other types of data 
which can be brought to bear upon the problem. Fur- 
thermore the evidence from chromosome morphology 
does not support adequately the conclusions which have 
been drawn. 
The presence of latent knob-forming points, invisible 
to the observer because adequate knob material is lack- 
ing, is questionable. If, when abundant knob material is 
supplied through hybridization, the knob-forming points 
still remain invisible, then another assumption—that the 
points have become temporarily inactive through disuse 
—is required (Longley, 16). 
There is some question, too, whether the gradients 
affecting knob position actually exist. There is no neces- 
sity for assuming a gradient in teosintes with terminal 
knobs; if there is a gradient for terminal knobs there may 
well be a similar one for chromosome ends and telomeres. 
The internal knobs of maize certainly follow a rough pat- 
tern, at least they are not distributed at random over the 
chromosomes. No knobs are found in the immediate vi- 
cinity of centromeres. Perhaps there is some significance 
in the fact that they are most frequent in the general re- 
gion where, if Darlington’s (6) figures of chiasmata in 
maize are a criterion, crossing-over should be at a maxi- 
mum. In any case the actual distribution of the knobs 
shows a poor agreement with the theoretical gradient 
(Longley, 14). A better fitting gradient could, no doubt, 
be calculated but perhaps the knob positions do not con- 
form to any gradient in the sense in which the term has 
been used. 
Still greater difficulties are encountered, however, when 
one attempts to carry to a logical conclusion, in terms of 
[ 241 ] 
