44 Referate. 
even if we admit this possibility, it would not assist the mutation theory 
as an evolution theory, because an evolution by repeated losses in incon- 
ceivable. The mutation theory requires progressive mutants, and “there is 
not a particle of proof for such an occurrence”. 
Mendel has shown us the manner in which new forms arise, viz. by 
hybridization, as, by crossing two monogametic individuals of different con- 
stitution, we get a polygametic hybrid. Hybrids are the only known source 
of the origin of new forms, some of which are heterozygotes, others of 
which are homozygotes, e. g. new genospecies; and in this respect it is 
indifferent if the parents are “pure species” or hybrids because the result 
has nothing to do with the origin of the gametes, but only with their 
constitution. 
The question: how does a new form perpetuate its constitution by 
transmitting it from parent to offspring — the conception which we call 
heredity —, is answered by Lotsy in an completely agnostic manner: We 
know “absolutely nothing” about heredity, and cytology does not learn us 
anything in this respect. Perhaps future researches into haploid organisms, 
especially mosses, will give us a key to understanding; but for the present 
we must stick to known facts and work with the origin of diploid organisms 
as far as possible. 
The fact that the individuals which we include in a species (Linneon) 
resemble each other more than they resemble other individuals and so to 
say simulate a genospecies, requires an explanation. 
In order to get at this, Lotsy calculates what will happen (1) if self- 
fertilisation follows a cross between two different organisms, and (2) if free 
intercrossing follows such a cross. In the first case we have Jennings’s 
calculation which tells us that in course of many generations nearly all 
individuals become homozygotes, as the relative number of heterozygotes 
diminishes and at last becomes minimal; thus a strictly selffertilising species 
consists of a number of different pure lines. In the second case the 
theoretical calculation (by Baur and Reimers) gives the result that the 
number of homozygotes and heterozygotes is about equal; but the real 
result is another, as is evident from the fact that species within which free 
intercrossing occurs, mostly have a rather uniform aspect. The cause to 
this phenomenon is the great role played by “intralinneontic” selection. 
Based upon interesting considerations and upon examples, Lotsy, therefore, 
concludes that a species (Linneon) is “a vestigial group of a once much 
larger group of differently constituted types, born from a cross, which is 
apt to simulate a species (genospecies) by the overwhelming majority of the 
dominant types it contains, as a result of free intercrossing, combined with 
a favoring of the dominant by a process of selection, weeding out the weaker 
or more conspicuous recessives; this uniformity being more apparent than real 
because pure dominants are indistinctible, in most cases, from dominant hybrids”. 
We find in nature that the species are, mostly, rather distinct, and 
the reason for this is that while free intercrossing within a species is the 
rule, it is the exception between different species, because such crossing 
is made difficult in different ways (aversion, differentiation in space, etc.). 
Our old species “are thus something more than mere conceptions of the 
human mind; they are natural intercrossing communities of differently con- 
stituted types’. They owe their origin to an occasional cross and their 
persistence to the bars which nature normally keeps closed to prevent 
crossing, but exceptionally opens. In badly limited species the normally 
