THE EXPERIMENTAL STUDY OF GENETIC RELATIONSHIPS 1 37 
by mutation, and by segregation from hybrids. Certain ultra- 
Mendelians maintain that only the latter method is operative, but their 
argument, which will be discussed after certain examples have been 
called to mind, can be supported by little or no evidence. As examples 
of recessive mutations, one naturally thinks first of the remarkable 
series of 150 which Morgan has observed to originate from the fruit 
fly, Drosophila ampelophila. All of them are recessive to the typical 
form. On the botanical side there is no correspondingly complex 
series of recessives known, of which all the members have originated 
under scientifically controlled conditions, but there is hardly the least 
doubt that the cases of the sweet pea (Lathyrus odoratiis) , of Primula 
ohconica and of Primula sinensis are quite comparable with that of 
Drosophila. Some members of each of these large series of forms are 
known to have come about as mutations, and probably all did. 
Bateson, who has devoted especial attention to the sweet pea, is sure 
that all the varieties are recessive to the wild prototype, Lathyrus 
odoratus. It seems sufficiently well proved that recent hybridization 
has not modified Lathyrus odoratus, for no other plant has been found 
which will cross with it. 
In nature there are many examples of species and varieties which 
bear simple Mendelian relationship to one another. Such a case has 
been discovered by Trow in the series of elementary species compre- 
hended under the name Senecio vulgaris. It also appears that in 
Antirrhinum, there are species which seem to be very unlike, neverthe- 
less all of the differences between them can be determined by Mendelian 
analysis. A large number of stable forms segregate from a hybrid 
between two such species, and these forms are themselves indistin- 
guishable from species. Their fertility is unimpaired. None of them 
contains a single character which is not identical with, or recessive to, 
a corresponding character in one or the other of the parents. Baur 
and Lotsy have studied the hybrids of Ajttirrhinum molle and A. majus 
from a Mendelian standpoint. The latter has been so impressed by 
the results that he has come to believe that there is no source of vari- 
ation in nature except hybridization followed by segregation ; it seems 
to him the sole method of species formation. He explains the so-called 
recessive mutations, no matter how rare they may be, by the assump- 
tion of previous hybridization and a sufficiently large number of 
multiple factors. It seems fair to ask the holder of this view how the 
forms originate which supply the characters to be assembled and 
