586 REPORT—1904. 
and each needs separate investigation. A remarkable case occurred in stocks, 
showing the need for caution in dealing with contradictory results. Hoary leaves 
and glabrous leaves are a pair of allelomorphic characters. When glabrous 
races were crossed with crossbreds, sometimes the results agreed with simple ex- 
pectation, while in other cases the offspring were all hoary when, in accordance 
with similar expectation, this should be impossible. By further experiment, how- 
ever, Miss Saunders has found that certain glabrous races crossed together give 
nothing but hoary heterozygotes, which completely elucidates such exceptions, 
There is every likelihood that wherever segregation occurs similar analysis will 
be successful, 
Speaking generally, in every case the first point to be worked out is the magni- 
tude of the character-units recognised by the critical cell-divisions of gameto- 
genesis, and the second is the specific consequence of all the possible combinations 
between them, When this has been done for a comprehensive series of types and 
characters, it will be time to attempt further generalisation, and perhaps to look 
for light on that fundamental physiological property, the power of cell-division, 
Segregation and Sex,—Acquaintance with Mendelian phenomena irresistibly 
suggests the question whether in ad/ cases of families composed of distinct types 
the distinctness may not be primarily due to gametic segregation, Of all such 
distinctions none is so universal or so widespread as that of sex: may it not be 
possible that sex is due to a segregation occurring between gametes, either male, 
female, or both ? It will be known to you that several naturalists have been led 
by various roads to incline to this view. We still await the proof of crucial 
experiments ; but without taking you over more familiar ground, it may be useful 
to show how the matter looks from our standpoint. As regards actual experi- 
ment, all results thus far are complicated by the occurrence of some sterility in the 
hybrid generation. Correns, fertilising 9 Bryonta dioica with pollen from 
& B. alba obtained offspring (F,) ether fg or 9, with only one doubtful excep- 
tion. Gartner found a similar result in Lychnis diurna 9 x 0 L. Flos-cuculias 2, 
but only raised six plants (4 g,2 9). From LZ, diurna 92 x § Silene nocti- 
flora as g he got only two plants, spoken of as females which developed occasional 
anthers. These results give a distinct suggestion that sex may be determined 
by differentiation among the male gametes, but satisfactory and direct proofs can 
only be obtained from some case where sterility does not ensue. 
Apart, however, from such decisive evidence—which, indeed, would be more 
satisfactory if relating to animals—several circumstances suggest that sex is a 
segregation-phenomenon. Professor Castle in a valuable essay has called attention 
to distinct evidence of disturbance in the heredity of certain moths (Aglia tau 
and dugens, Standfuss’s experiments; Tephrosia, experiments of Bacot and others, 
summarised by Tutt),! where the disturbance is pretty certainly connected with 
sexual differentiation, Mr, Punnett and I are finding suggestions of the same thing 
in certain poultry cases. Mr, Doncaster has pointed out that the evidence of 
Mr. Raynor clearly indicates that a certain variety of Abraxas grossulariata, 
usually peculiar to the female, is a Mendelian recessive. It is scarcely doubtful 
that this will be shown to hold also for some other female varieties, e.g., Colias 
edusa, var, helice, &c. We can therefore feel no doubt that there is some entangle- 
ment between sex and gametically segregable characters, A curious instance of a 
comparable nature is that of the Cinnamon canary (Norduijn, &c.), and similar 
complications are alleged as regards the descent of colour-blindness and 
heemophilia. 
‘In one remarkable group of facts we come very near to the phenomenon of sex. 
Experiments made in conjunction with Mr. R. P. Gregory have shown that the 
familiar heterostylism of Primula is a phenomenon of Mendelian segregation. 
Short style, or ‘thrum,’ is a dominant—with a complication;* long style, or 
‘pin,’ is recessive; while equal, or ‘ homostyle,’ is recessive to both, 
1 Trans. Ent. Soc. Lond., 1898. 
* It is doubtful if ‘thrum’ ever breeds true, as both the other types can do, 
Perhaps ‘thrum’ is a Halbrasse of De Vries, 
