THE EXPERIMENTAL STUDY OF GENETIC RELATIONSHIPS 1 53 
mutation occurred. We are not told of any check cultures of sister 
plants from untreated ovaries, or of any further test of the strain for 
mutability. MacDougal referred to this experiment in 191 1 as the 
most conclusive which he had yet performed, and stated that the 
mutation had come true from seed for five generations. There is no 
doubt that it is a very interesting mutation, especially in view of the 
fact that its behavior on crossing with the parent strain resembles that 
of Oe. gigas. The evidence that it was induced by the zinc sulphate 
solution is, however, quite insufficient. In the case of Raimannia 
odorata 13 mutations are said to have been obtained from ovaries 
which were injected in 1905. We are not told whether the strain 
had been previously purified by self-pollination or not. One ovary, 
injected with 10 per cent, sugar solution, gave two mutations; another, 
injected with i : 1,000 calcium nitrate solution, gave 10 mutations; 
a third, after exposure to radium, gave one mutation. We are not 
told whether the injected ovaries were all on one plant or not. There 
s no record of any check experiments with untreated ovaries, nor any 
record of the size of any of the cultures. More injections were made 
in 1906, with the contradictory result that calcium nitrate, which had 
been so potent the year before, induced no mutations at all. Three 
mutations were found in the progeny from "capsules" treated with 
I : 2,000 zinc sulphate, and also ''other combinations" which were not 
followed in subsequent years. We are told, moreover, that the effect 
of the first injections with calcium nitrate persisted in following gener- 
ations, for normal plants belonging to the strain from the ovary which 
was treated in 1905 with calcium nitrate gave rise to the same mutation 
which was supposed to have been induced by the chemical treatment. 
It seems to me that these facts can only mean that the strain used by 
MacDougal was in a highly mutable condition and that the experiments 
were not properly checked. It is especially noteworthy that Compton, 
also working with Raimannia odorata, has been unable to induce any 
mutations by the use of MacDougal's method. His strain was doubt- 
less a non-mutable one. It is much to be hoped that more decisive 
results will attend future work along this line, and that whoever under- 
takes such experiments will adequately check them. 
The third aim of the experimentalist was to provide a basis from 
which the systematist might determine deductively the genetic re- 
lationship of organisms. It must be admitted that except in the case of 
Mendelian varieties little progress has been made. Nevertheless 
