490 TRANSACTIONS OF SECTION D. x 
Mendelians that characters transmitted as units must have arisen as units, and 
it 1s certain that Mendelism has shown how loss of characters and new com- 
binations produce new varieties or types. It is reasonable to conclude from 
present knowledge that non-useful diagnostic characters have arisen as the 
result of gametogenesis and conjugation; but the principles of Mendelism or 
mutation are not applicable to the phenomena of adaptation. 
In the first place when we see, as in the frog, the flat-fish, or the caterpillar, 
adaptation to two quite different sets of conditions in the individual life, it is 
impossible to believe that such transformation was due to mutations not caused 
by the external conditions. There is no evidence that the necessary gradual 
changes could occur unless the conditions produced them; if so, why have they 
not occurred in other cases when the conditions were absent. 
In the second place we have the phenomena of secondary sexual characters, 
of which one of the most impressive and most fully investigated is that of the 
antlers of stags. The Mendelian merely regards such characters as mutations 
which are coupled with primary sex. But primary sex is determined at 
fertilisation, and such secondary sex characters have been shown to be dependent 
on the presence and function of the gonads. Characters which are determined 
in the gametes are not generally affected by computations of gonads at any 
part of the body in after life. It has been shown that the effects of castration 
on the development of secondary sexual characters are due to the stimulus of 
chemical substances produced by the gonads, especially in their functional 
activity. 
No hypothesis explains these facts except the Lamarckian, namely, that the 
stimuli involved in the use of the organ originally produced them by causing 
hypertrophy in the part of the soma affected, and that in course of generation 
the tendency to this hypertrophy was transmitted to the gametes. The hormone 
theory explains how such transmission may be effected. The hypertrophied part 
gives off chemical substances or hormones which circulate through the body, 
and acting on the gametes stimulate those parts of them which are destined to 
develop the same parts in the next generation. The transmitted effect may be 
infinitesimal at first, but if continued for many generations would account for 
the phenomena we now observe. 
This, of course, would account for the transmission of all somatic modifica- 
tions due to external stimuli, and a special application of the theory is needed 
to explain the peculiarities of functional secondary sexual characters. 
In the first place the stimuli in these cases have acted only on individuals of 
one sex, on the males in stags, on the females in the case of the mammary 
glands. On any other theory a variation occurring in one sex would be inherited 
by both sexes unless it was coupled with primary sex, and then it would be 
wanting in the other sex. But antlers are not wanting in females nor mammary 
glands in males: they are only not developed. On the hormone theory the 
somatic modifications were produced at the time when the gonads were giving 
off their hormones, and thus the tendency which is inherited is to develop these 
modifications in the presence of those hormones and not otherwise. Then we 
can understand why the organs develop only at puberty, and often only develop 
during the period of sexual activity, being shed or absorbed at the end of that 
period and re-developed. 
6. Some Facts regarding the Anatomy of the Genus Pegasus. 
By Professor Hector F. HE. Junaursen. 
The facts, briefly condensed in the following abstract, have—for the greater 
part—hitherto been overlooked or unknown. 
Cranial Skeleton.—Opisthotics, alisphenoids, orbitosphenoids, and _basi- 
sphenoid absent ; no eye-muscle canal. Posttemporal (suprascapular) forms part 
of the skull. Three stout infraorbitals, the middle and posterior firmly con- 
nected with the preopercle. Opercular apparatus complete. The large flat 
preopercle, covering most of the lower face of the head, has generally been 
taken as ‘homologous to operculum, prexoperculum, and _ suboperculum’ 
(Giimther), while the very small opercle and subopercle, hidden in thick skin, 
have completely escaped attention. Interoperculum slender, widely separated 
