SPONTANEOUS CROSSING IN THE GARDEN BEAN 399 

Of ail investigated varieties the chocolate brown beans have given 
the greatest percentage of hybrids, and this has been the case also 
in 1918 and 1919; it is not very probable that the result is a mere 
chance. The greatest percentage of crossings ever found in any 
isolation was in such a chocolate brown bean. There were in this 
case 5 plants with a progeny of 34:individuals in all, 6 of them 
(17,7 %) being hybrids. Only 2 of the plants, however, had been 
cross-pollinated, but about half of their progeny (46,9 %) were hybrids. 
The values of the spontaneous crossing obtained by using the 
above mentioned experimental method may scarcely be considered 
as the maximal ones, but they may yet be very near those. In order 
to obtain such maximal values it had surely been more convenient 
to leave only one plant of the variety to be tested. Cross-pollination 
between plants of the same isolation had then been avoided. In this 
case it should be necessary to sow several seeds in each isolation; 
immediately before the flowering a thinning out should take place so 
that only one single plant was left. I had no time to make such 
a thinning out, however, on account of other theoretical and practical 
investigations that were under way. 
Another procedure is to sow one single seed of the variety to be 
tested in isolations located in sufficient large distances from each 
other (not in the cultures of beans) and to sow around it seeds from 
the other variety. The advantage of this method lies in the fact that 
it then is possible to compare the tendency of cross-pollination of two 
varieties with greater certainty than if the isolations are sown in the 
large bean-cultures, where generally many varieties are cultivated, and 
consequently the risk is at hand that the insown stock could be cros- 
sed with other plants than the surrounding. This method may be 
altered in such a manner that only one plant of each variety is 
allowed to remain in each isolation. In this case, however, the risk 
is close at hand that the visits of insects would be too-few, resulting 
in a wrong interpretation of the data obtained. 
The percentage of spontaneous crossings are in some cases rela- 
tively considerable; it would, therefore, be best to isolate the plants 
to be used for theoretical investigations. Special caution should be 
exercised in the judging of segregations which do not seem to be in 
correspondence whith the Mendelian ratios if the plants have not been 
isolated. A still more rigid criticism is required when supposed muta- 
tions are found. It becomes necessary to investigate whether these are 
