1466 THE BUTTERFLIES OF NEW ENGLAND. 



eyes turn pink, the next day inky ; two days later the wings have turned 

 of an inky hue, and the last day the whole body. In Mr. Lintner's expe- 

 rience, the whole period in August was only six or seven days ; and in the 

 only instance yet known ofexperiment with a hibernating caterpillar, by Mr. 

 Edwards, the chrysalis period in spring was between six and seven weeks. 



Life history. There are three apparitions of this butterfly annually ; 

 for it is partly single, partly double, possibly partly triple, brooded, but 

 always passes the winter in the larval condition. The first butterflies make 

 their appearance early in May and continue into June, and occasionally in- 

 to the first week of July. They lay their eggs at once and continue to do 

 so in the first, rarely in the latter, half of June. The eggs are hatched in 

 about ten days, and the earliest larvae are nearly full grown by the middle 

 of June ; the majority change to chrysalis in the latter half of June, and 

 the first week in July ; they pass nearly or quite a fortnight in this condition, 

 and the earliest of the second brood of butterflies makes its advent about 

 the middle of July, though the males sometimes appear as early as the 6th. 

 It, however, happens that all full fed caterpillars of this date do not change 

 at once to chrysalis, or perhaps even the major part of them ; others re- 

 main immovable through the entire hot season, as well as often through the 

 winter ; thus of three specimens reared together, which closed up their cells 

 in the first four days of July, one changed at once to chrysalis (July 3) and 

 emei'ged July 14, the two others remained caterpillars in closed cells with- 

 out feeding and without changing to chrysalis all through July and Au- 

 gust ; in September one of them changed to chrysalis, and emerged at the 

 end of the month, the other was still a caterpillar on October 1 and so 

 passed the winter. Mr. Saunders also reared, in September, an imago 

 from a caterpillar of the first brood. To return, however, to our second 

 brood. It continues on the wing until the middle of August, occasionally 

 nearly to the end ; eggs are laid all through July, and probably into August. 



The third brood is plainly a mixed one, made up partly by direct and 

 undelayed descent from the second brood (for Mr. Lintner has raised 

 the butterflies directly after six and seven days in chrysalis) partly by 

 descent from the lethargic caterpillars of the first brood (for, as stated 

 above, one of my caterpillars which was full fed early in July changed to 

 chi-ysalis in September, and gave the imago the same month). This 

 brood appears about the middle of August, and continues to emerge 

 through September. Eggs are laid in August and at least early in Sep- 

 tember ; but whether any but the August caterpillars reach maturity before 

 winter is not known, nor whether they perish if they do not become full 

 grown. All hibernating caterpillars that we know of in the whole genus 

 do all their eating before winter and eat nothing in the spring, but change 

 early to chrysalis in their hibernacula ; this was the case with one of this 

 species raised by Mr. Edwards in AVest Virginia, which appeared as butter- 



