260 POLYMORPHIC BUTTERFLY. 
pearance of a subsequent brood, but usually until the next spring, 
The spring brood (Walshii-Telamonides) is therefore by no means 
wholly produced from ¢hrysalides of the final brood of Marcellus, 
but in large measure from those of all the earlier broods, even in- 
cluding the earliest Walshii; the proportion of chrysalides which 
continue until spring increases asthe season advances, Mr. Edwards — 
statements showing that of those produced from eggs laid in April, 
more than ten per cent. pass over, those from eggs laid May 1-25 
about thirty-five per cent., from that time until the end of June 
from fifty to sixty per cent. and from those laid in July about sev- 
enty per cent. Walshii and Telamonides, then, produce Marce us 
the same season, or either Walshii or Telamonides in the spring; 
Marcellus produces itself the same season, or one of the’ others 
in the spring; but neither Walshii nor Telamonides is produced 
the same season by any of. the varieties.* : 
We will now consider the life-history of Ajax. The insect Is 
multibrooded and winters as a chrysalis. The earliest variety) 
Walshii, ‘‘appears in the Kanawha valley (W. Va.), from the 15th 
to 20th of March, by which time the peach trees are usually 
blossom. On these. the females may certainly be found, and $ 
little later, on the apple and in great numbers on the wild ph 
The males appear a few days earlier [than the females] and are 
be seen by the water-side or upon the road, but rarely upon flow 
The larvæ feed on the papaw, and as this is one of the latest 
of our trees to put forth its leaves, the butterflies are out at least 
from two to three weeks, before the young shoots of the food plant 
are visible. But no sooner do these appear than the fem 
hasten to deposit their eggs.” This is early in April and the 
* Th orded ptions to this rule (which serye only to strengthen it) are! sal 
: a Tel id ptured, September 12th, and a Marcellus April Hth; 
other í tions oceurred in the course of Mr. Edwards’ experiments. 
igable worker has traced the history of more than two hundred individuals from 
to butterfly, and of the: ly two did not follow the usual course; these 7 
to a brood produced from eggs laid by Walshii before the middle of April; pr sa 
va 1. n C E P T E N" e PE éb nzr ting a pair pik 
lo 
LOW 
Marcell ) hatched bet June Ist and 6th; fifty-eight chrysalides produ 
lus, one Walshii and one Telamonides; so that the solitary specimen of Walshil 
far out of season, for Mr. Edwards expressly says: “about the first of June 
i u 
bame season. It must be remembered that similar instances of 
by no means very uncommon among butterflies. 
t Edwards, Butterflies of N. America, Unless otherwise stated, all these Fe 
are based upon Mr. Edwards’ observations in W. Virginia. Ja 
