NYMPHALINAE: ARGYNNIS ATLANTIS. 577 



80011 as hatclicd and before eating. It undergoes its later transformations 

 in the spring, probably awakening at an early period, as the larva feeds 

 upon violets, which are early in leaf. In the White Mountains, the but- 

 terfly appears about the middle or latter part of June, but does not 

 become conunon until the close of the month. In the Catskills, according 

 to Mr. Edwards, the males appear early in June. It is the earliest of the 

 species of Argynnis in any given locality. The butterfly continues abund- 

 ant until the middle of August and is still upon the wing in September. 

 Fernald is certainly mistaken in saying that in jNIaine it "is on the wing 

 from the middle of June to the early part of July" only. The eggs are 

 laid during the latter part of August, when the numbers of the butter- 

 flies begin to diminish. Miss Soule sent me some laid in Stow, Vt., on 

 August 22. Mr. W. II. Edwards obtained them in the Catskills August 

 28-31, and all I have had were laid at about the same time. I have 

 opened old and worn females at the White Mountains the middle of 

 August, and found the eggs only partly developed ; later, on August 19, 

 I have found them very nearly ready to be laid. The eggs hatch in a lit- 

 tle more than a fortnight* and the caterpillars go immediately into winter 

 quarters. Of some which Mr. Edwards kept over the winter on ice and 

 brought out again in West Virginia, March 21, one moulted April 15, 

 the second moult was passed by different caterpillars April 25-29 ; after 

 which one caterpillar, the only one carried to chrysalis, moulted success- 

 ively, on May 9, 19 and 25, and changed to pupa June 3. The length 

 of the several larval stages, not counting the wintering one, was 10, 14, 

 10, 6, and 9 days. The chrysalids attach themselves to the under side of 

 logs on the ground and in similar places. 



Habits, etc. The butterfly is wilder and more difficult to capture than 

 the allied New England species, which it resembles closely in its flight. 



The male has a very perceptible odor which I found it difficult to 

 determine ; it seemed to me at first to have somewhat the fragrance of the 

 pine woods ; but on applying to Miss Soule, she at once said it was the 

 odor of sandal- wood, which exactly describes it. She added that speci- 

 mens she had brought into her room at the mountain hotel made it seem 

 close, and a friend who called inquired on entering : "Why ! do you bring 

 your sandal-wood box up here ?" The odor is apparently equally strong in 

 new and old specimens, and arises from the fringed androconia as I have 

 found by careful tests. It is often retained for months after death by 

 specimens enclosed in papers. 



The male follows the female closely, even long before oviposition. The 

 female, when wearied with flight from its pertinacious mate, will rest with 

 erect wings motionless upon the ground ; its mate immediately takes posi- 

 tion beside it, distant about half an inch or an inch on one side and behind 



*Soiiic laid August 21-22 and sent to "West Virginia hatched a little sooner, September 3. 



73 



