286 
Letters to the Editor. 
[Zhe Editor does not hold himself responsible for 
opinions expressed by his correspondents. Neither 
can he undertake to return, nor to correspond with 
the writers of, rejected manuscripts intended for 
this or any other part of NATURE. No notice is 
taken of anonymous communications. | 
The Function of Mendelian Genes. 
In Nature for January 20, p. 74, Prof. MacBride 
makes a statement which appears to me to rest upon 
a fallacy. Since this fallacy is not uncommon, and 
since it concerns a very fundamental problem, I feel 
that perhaps a discussion of it in these columns may 
serve a useful purpose. 
Prof. MacBride, in the review referred to, writes as 
follows: ‘' Prof. Reinke encounters the Mendelian 
“gene ’ and in our opinion takes it far too seriously. 
. . It is becoming every day clearer that a‘ gene’ is 
not a definite unit of structure at all, but simply the 
measure of the amount of pathological damage which the 
hereditary substance has undergone. (Italics in the 
original.| It is a measure, in a word, of the ‘ im- 
perfection of regulation.’ ”’ 
The fallacy involved is simply this—that Prof. 
MacBride is using the word ‘‘ gene”’ as if it meant 
“mutant gene.’’ A mutant gene is, strictly speaking, 
that portion of the hereditary constitution which is 
responsible for the characters of a mutation observed 
to arise in Nature or in the course of experiment. 
I take it, however, that Prof. MacBride is considering 
all definite variations which are inherited according 
to Mendel’s laws, whether their origin was observed 
or no; this at any rate is now a legitimate extension, 
and I shall employ ‘“‘ mutant gene’”’ to denote the 
altered portion of the hereditary constitution respons- 
ible for variations inherited in a Mendelian way. 
Now in point of fact, as Morgan himself and other 
writers have taken great pains to point out, the 
discovery of each new Mendelising variation, of each 
new mutant gene, implies also the discovery of an 
allelomorphic ‘‘ normal gene”’ responsible for the 
production of the ‘‘ normal”’ structure and function 
of the part or parts affected by the mutation. 
The work on Drosophila has completely proved that 
Mendelian genes are carried in the chromosomes. 
Prof. Bateson, for long sceptical on this point, finally 
conceded it last year after seeing the work of Morgan 
and his pupils at Columbia University, New York. 
Further, it has proved that within each chromosome 
the genes are arranged in a definite way ; the observed 
facts are intelligible if we assume that the genes are 
arranged in a constant and linear order, while no satis- 
factory alternative hypothesis has been put forward. 
In any case, the order is identical for all the 
homologous chromosomes of the species which are 
tested. A mutant gene, therefore, occupies a similar 
position in one particular chromosome to that which 
1s occupied in the corresponding chromosome of the 
normal wild strain by an allelomorphic gene which 
has not mutated. The mutant differs from the non- 
mutant gene by some definite alteration, presumably 
of a chemical nature: the existence of series of 
multiple allelomorphs, together with other evidence, 
proves that a recessive gene is not a mere total 
absence of the something we call the dominant gene. 
That is to say, in the chromosomes of the normal 
wild-type animal or plant there exist a large number 
of genes or factors the chemical constitution of which 
cannot be altered without giving rise to ‘‘ mutations.”’ 
Some of these mutations are pathological; others 
(pace Prof. MacBride !) are not.t 
? See Science Progress for 1921, where Prof. MacBride, in answer to two 
letters of mine, eventually admitted that not all were pathological. This 
question, however, does not concern the present argument at all. 
NO, 2783, VOL. 111] 
NATURE 

[Marcu 3, 1923 

But even if they were all pathological, this would 
not alter in the very slightest the fact that their non- 
mutant allelomorphs constitute an orderly series of 
discrete units distributed in heredity by the chromo- 
some mechanism (i.e. according to the laws of Mendel 
as extended by later research), and all necessary for 
the noymal development of the individuals of the 
species. Of course, if all mutatigns were pathological, 
Mendelian genes would have no significance for 
evolution. However, they would even so continue to 
have the most fundamental significance for normal 
heredity. 
Perhaps my meaning may be made clearer by a 
brief example. In the wild house-mouse, each hair 
is black with a yellow band across it, the yellow and 
black blending to give an appearance of grey; grey 
of this type is technically called ‘‘ agouti.”” Black 
mice are mutants in which the yellow band is absent ; 
this condition is recessive to normal. Yellow mice, 
on the other hand, have the yellow pigment extending 
the whole length of the hair ; and yellow is dominant 
to grey (agouti). It is also dominant to black. The 
three types of colour and their behaviour in crosses 
can only be explained if we suppose that there is a 
definite gene responsible for the production of yellow 
pigment, and that this exists in three separate states 
—a “ strong ”’ state when a great deal of the pigment 
is produced, a “‘ medium ”’ state in which a moderate 
amount is produced, and a “‘ weak ”’ or non-effective 
state in which no yellow pigment is formed at all. 
The gene in its medium state is responsible for the 
particular proportion of yellow which we see in the 
hair of wild house-mice. The three states are all 
allelomorphic to each other, the “‘normal’”’ being a 
Mendelian recessive as against yellow, a Mendelian 
dominant as against black. It is impossible to escape 
from the idea of a discrete unit of definite composition 
helping to determine coat-colour in the normal 
animal, strictly comparable to the homologous units 
responsible for the two “ mutations.” 
This example is also of service as regards the 
abnormality or otherwise of mutations and mutant 
genes. The alteration productive of all-yellow is 
decidedly pathological. Even a single dose of this 
gene leads to excessive fatness, and two doses cause 
death of the foetus zn utevo. The recessive “ black ” 
gene, on the other hand, does not appear to be 
responsible for any pathological effects. What is 
more, there is no evidence against the view that this 
mutant black is strictly comparable to the black of 
melanistic mammals in Nature, and the similarity 
is so great that the onus of proof lies on those who 
would dispute this homology. 
In any event, there are two quite distinct aspects 
of the gene question—the genetic or hereditary, 
and the evolutionary. Mutations in Mendelian genes 
may or may not have been responsible for variation 
which has played a part in evolution. This I do not 
propose to discuss here, except to say that I know 
from many conversations that Prof. MacBride’s views 
are too sweeping for a number of zoologists. But as 
regards inheritance within the species, the Mendelian 
gene—once the fallacy of confusing ‘“‘ mutant gene”’ 
with 
obviously of importance. 
We are to-day in a position to make a calculation of 
the order of magnitude of the number of genes in the 
chromosomes of Drosophila. It is certainly more than 
1000 ; probably more than 2000; certainly less than 
20,000. The effects of alterations in more than 200 of 
these genes (i.e. mutations) have been observed and 
studied ; there is no sign that the rapid stream of new- 
discovered genes is slackening. With such a number 
of genes responsible for keeping the development of a 
little fly in the straight and narrow path of normality, 
“gene’’ is seen and avoided—is clearly and 
a 
Com 
