300 
identical genes in the common ances- 
tor of the two 13-chromosome types.* 
III 
There may seem to be a contradic- 
tion in what has been stated. I have 
argued that related species have essen- 
tially the same sets of genes, and yet I 
have admitted that within one species 
there may be a series of different but 
similar genes at any one locus in differ- 
ent members of a population. If one 
studies the characters conditioned by 
the various genes at any one locus they 
turn out to be related. The impres- 
sion is that all the genes at any one 
locus are developmentally alike; ap- 
parently they are carrying on the same 
function, but with varying degrees of 
efficiency—often with different effi- 
ciencies (not necessarily parallel) in 
different parts of the organism. If the 
genes are thought of as being catalysts, 
or, as is perhaps more probable, as 
conditioning the presence of specific 
catalysts, then one may make a rough 
analogy to a lock-and-key system in 
which the different genes at any one 
locus are keys to the same lock, but do 
not all fit it equally well. It is in this 
sense that one may conclude that 
related species have like genes—per- 
haps not identical, but certainly very 
similar, and carrying out the same 
functions. They are keys that fit the 
same lock. 
The chemical composition of the 
genes is not accurately known, but 
evidence is accumulating that seems to 
make it most probable that they are 
largely or entirely nucleoprotein in 
nature. The indication of a protein 
component is of considerable interest, 
for it suggests that the problem of gene 
4 This account is somewhat simplified. In 
particular, some of the evidence is derived 
from comparisons between different 13- 
chromosome types. There is, however, a pos- 
sibility that these are themselves derived from 
doubling of the chromosomes of still earlier 
hybrids between forms with lower numbers. 
So long as this possibility exists these species 
cannot furnish conclusive evidence for a 
change in function of genes. 
ANNUAL REPORT SMITHSONIAN INSTITUTION, 1948 
specificity—that is, how it happens 
that each gene has properties different 
from those of its fellows—is an aspect 
of the problem of protein specificity. 
This in turn means that the geneticist 
may expect help from the techniques 
and results of other biologists who are 
concerned with protein specificities— 
such groups as enzyme chemists and 
immunologists. 
There is a current tendency to look 
upon protein specificities as being due 
to the way in which the molecules are 
folded—to their shapes rather than to 
their gross chemical composition (6). 
On this basis the primary specificity 
may be supposed to be that of the 
genes, the other proteins having their 
specific properties impressed upon 
them by the genes (3). This is sug- 
gested by two circumstances; it is clear 
that genes can in some way impress 
their specificities on new material, for 
this is what must happen each time a 
gene reproduces itself at cell division, 
and it is also clear that antibody pro- 
teins somehow have their specificities 
impressed on them by the correspond- 
ing antigens in cases of acquired im- 
munity (5). It thus seems probable 
that the lock-and-key analogy sug- 
gested above is a valid one, though one 
must think of the keys as reproducing 
themselves without the intervention of 
a locksmith. 
Every cell of an organism has, in 
general, the same set of genes as the 
other cells in the same individual. 
The problem that arises at once is: 
how is differentiation possible? If the 
characters of an organism are depend- 
ent on its genes, it follows that the 
different properties of the various parts 
of one organism are under gene con- 
trol, yet the same genes are present in 
all these parts. This may mean, in 
terms of the lock-and-key interpre- 
tation, that in any given part of the 
organism all the keys are present but 
only some of the locks; in other words, 
that the outcome is determined by 
which substrates are present rather 
than which gene-controlled enzymes. 
This is, however, probably an incom- 
