29 
for as follows: Inbred from captured Js prior to my 1894 paper; 
one short series in-bred from these : three broods of which none have yet 
emerged from pupa ; and two of which 1 did not keep suliiciently exact 
data as to numbers, one being a batch of some BO or more black speci¬ 
mens bred from a captured black 5 , and the other a batch of similar 
size (and likewise all black) from a of these paired with a 5 from 
another totally black strain. Jlesides all these, I once kept three or four 
black 5 s together, obtained a large number of ova, and again reared 
})lavJx .sj/cciiiK'iis Old;/. 
It will suffice to say of these early breedings, that the result tallied in 
the main with those of later experiments. About 10 captured black 2 R 
yielded only black offspring, hence were presumably fertilised by black 
(?s; one captured red J yielded red only (16 in number); another, 
red and intermediate only (but the total number was very small) ; 
these two may have been fertilised by red s ; and four others gave 
red and hlack in about equal proportions ; these were probably fertilised 
by black ^ s. 
The starting-point of my present work in pedigree-breeding is April 
1895, and the first material was furnished by the offsping of three J s 
captured at Sandown in August, 1894—one red and two black. I was 
thus able to pair black with black, red with black (reciprocally), and 
red with red, though the last-named pairing necessitated inbreeding 
from my single red strain. Further material has been added from 
time to time, always from ? s captured at Sandown. I will not weary 
you with details of each of the BB experiments separately: but prefer 
to take a general survey of the results. 
J!(‘d fftii(d(‘K cajitiin'il .—Seven of these have given a rougii average 
of 54% red, 45% black ; red ? s lead in point of numbers, then black 
A s, then red S s, a,od lastly, black 2 s. It is an interesting coin¬ 
cidence that the above percentage corresponds irri/ clasdi/ with the 
average struck between those furnished by red ? s x red s, and red 
2 X black S • 
/iVd ft’iiialr fertilised hij hlaek male .—A large brood resulting shows 
B9% red, 61% black. 
lied females fertilised hi/ red males .—Nine of these show an average 
of about 68A% red, B0% black, the rest intermediate ; red 3 s come first 
numerically, then red 2 s, then black 3 s, and lastly, black 2 s. 
Jilaek females ea/itiired .—Four of these might almost be said to 
yield 100% black ; but as one brood (labelled X in my box) contains 
some specimens with a reddish blush, 1 have reckoned them as inter¬ 
mediates, and will give the percentage as 90i% black, 34 intermediate. 
Jllaek females fertilised hi/ red males .—Four of these give about 
46^-% red, 51% black, 2|% intermediate ; black 3 slead, then red 2 s, 
then black ? s, and lastly, red 3 s. 
lUaek females fertilised h// hlaek males .—Seven of these yield 9B-i% 
black, none red, 1-^-% intermediate. In this connection it should be 
noted that six of these seven broods had red $ grandparent, and in one 
case both parents were the offspring of a red 2 ; yet the “ li% ” really 
represents only a sini/le intermediate specimen. True, none of the 
broods were very large, but two of them contained 16 .specimens 
apiece, without a single intermediate. One of these two had a strong 
reddish tendeneij in the 2 parentage (the brood “ x ” already referred 
