51 
species which are to emerge in the summer ; I have no exact records, 
hut believe the period is about a week. 
Imago. —This has already been described above, and I am compelled 
to omit from this already lengthy paper, a discussion of its variation, 
and named forms. I have, therefore, only to deal with its time of 
appearance and its habits, and then to conclude with my notes on the 
distribution of the species. 
Time of appearance.— The following dates deal with the appearance 
of F. antmmiata in a state of nature: September-October (F. B. White, 
Sir T. Moncreiffe); late September-second week of November (Gregson); 
October 2nd, 1898(Allen); September 25th-()ctober 23rd, 1897 (Horne); 
end of September-early October (Christy—at Rannoch); November 
(Guenee) ; end of August-October (Aurivillius, Schoyen, Sec.) ; Sep¬ 
tember 18th-October (Sparre-Schneider); September 16th, 1890 
(Bohatsch); October 24th, 1897 (Ileyne). 
In captivity, this species emerges very erratically, and over a 
considerable period of time ; and when the larvie are forced at all, it 
will sometimes commence appearing very early. Three of my 1898 
specimens (which, as I have already shown, were very forward in the 
larval state) took advantage of some cold weather in the beginning of 
July to put in an appearance ; unfortunately I was not expecting them, 
and cannot give the exact dates. I have, however, exact dates for all 
the other specimens I bred, and Mr. Allen has very obligingly furnished 
me with those of his Enniskillen examples. I give the following 
records: My Aberdeen brood, July (before the 10th)—August 26th; 
one Kincardine brood, July (before the 14th)—October 3rd ; another 
Kincardine brood, August lst-October 28th ; Mr. Allen’s Enniskillen 
series, September 24tli-November 6th; Mr. Clark’s Rannoch brood, 
1895 (Fut. Jiec., vii., p. 289) August 5th-November lltli ; Schneider 
from Sydvaranger larvae, 1893 (Tromsu Mas. Aamh., xv., p. 79), cir. 
August 20th-September 6th, as against cir. September 20th-October, 
in a state of nature. All my available records go to show that the 
emerging period of diliitata is, on the whole, more regular and less 
protracted. 
Habits. —Of these I know next to nothing. According to Weaver, 
the typical form rests on the branches of birch, while the var. appro.vi- 
maria is found in fir plantations, and often at rest upon the trunks. 
Some of my correspondents have obtained specimens by beating trees, 
and these specimens will presumably have been resting amongst the 
branches. Mr. Gregson used to find them on tree-trunks in Delamere 
Forest, and he has an interesting observation (Fool., 1859, p. 634S) 
that “ the habits of the October and November insects are so different 
that anyone can tell which species he is about to take: 0. diliitata 
flies off the tree bole rapidly and generally upwards, 0. autumnata is 
one of the most sluggish flyers; an old friend once observed to me 
‘ nay, do not hurry, autmnnata will not go fast or far. ” Possibly, 
however, my friend Mr. Allen may have something to say as to the 
last part of this quotation ; I gather from one of his letters that he 
had at least one good chase after a specimen at Enniskillen. 
Distribution.— This is at present very imperfectly known, and I 
have had to work up my list almost exclusively from specimens I have 
myself seen ; the following includes all localities as yet known to me : 
Great Britain and Ireland ; Swansea (one specimen from Major Boberlsen), 
