80 
to), besides predominant red on the side (<? was one of a brood with 
22 red to 18 black). 
IntcrmtuJiati' fciindc fcrtiliwil Ini hhirl,- male. —One of these has 
produced as yet three black specimens only, but others of the brood 
are still in pupa. I ought to add here that some of my statistics above 
are worked out with the inclusion of broods not yet complete, but the 
specimens yet to emerge cannot materially modify the results. 
Summing up, I find that in black x red, or red x black, black is 
inclined to preponderate slightly (nearly as 4:8). The dominance of 
red in the seven broods from red captured $ s may well be accounted 
for, as hinted above, by fertilisation by red S' in some cases. In one, 
at least, labelled “ P ” in my drawer, I very strongly suspect this, both 
from the proportion (39 red to eight black, or H8°/o) and the character 
of the red specimens therein (especially compare the forms with those 
in brood ]\IRT, known to have both parents red). In red x red, red 
strongly predominates, nearly in jiroportion of 7 : 3 ; the statistics 
given in our Transactions for 1897, p. 18, shoAV that continued red 
selections steadily increase the percentage, and possibly in another 
generation or so a purely red brood could be obtained, though (if my 
present theories are correct) a reversion to the black type would always 
be liable to occur. In black x black the I’ed seems almost entirely 
Ainable to assert itself, and one almost Avonders hoAV the red A-ariety 
eA’er originated. 
I haA'e been A’cry much impressed by the general rcri/ direct response 
to immediate parentage, especially if black ; black x i)lack producing 
black only (irrespective of ancestry), red x red producing OA’er red, 
red X black roughly half-and-half, or black slightly in the ascendant. 
Possibly, hoAvever, if I had carried iny experiments further, I should 
have obtained more indications of reversion. The potency of the S 
and of the 7 parent seems roughly equal, nor does it, so far as my 
statistics shoAV, exert its influence more strongly on one sex than on 
the other in the progeny. 
Another thing Avhicli 1 have been much interested in observing is 
the occurrence of certain characteristic forms in certain broods. I 
noticed tAvo instances in my 1894 paper, as Avell as some others in the 
allied s/milicearia tferra/iaria) {Trans. Citi/ Lond. I'.'nt. & Ak If. Soc., 
1894, pp. 27 and 28), but noAv that I haA’e so much more material 1 
am able to point to many others. For instance, there is one pretty 
constant red form prevailing in my brood LMa, and (mainly) in 
L]\f, ML, and M, all of Avhich OAve their red element to the captured $ 
M. In brood P, on the other hand, the Avhitish outer area is A'ery 
striking, and quite different from anything found throughout the afore¬ 
mentioned broods, Avhile in brood Q the strong ochreoas tint in outer 
area is almost equally striking; these tAvo broods, and brood (), Avith 
its generally broad central band, Avere all emerging at the same time, 
and I Avas amused to notice hoAV easily I could ])ick out specimens of 
each brood on the setting-boards Avitliout reference to their labels. .\ 
broad central band is also characteristic of brood U, Avhich Avas also 
easy to distinguish from its dull-coloured contemporary The black 
form Avitli rod costa, Avhich I Avill call ab. rnjicostata, and Avhich I had 
only noticed to appeal' sporadically Avhen 1 Avrote in 1894 {l.r., p. 28), 
has appeared in almost every specimen of brood X,the red tinge some¬ 
times extending to other parts of the central band. But the broods 
