INSECT PESTS OE PINE AND! SPRUCE CONES: CXXXIX 
difficult to know whether these data have a general applicability; but it 
seems probable that they fairly; reflect the relative importance of the para- 
sites. The importance of these is considerably increased by their occurring 
together; and in 31.7 &Z of the localities more than 30 2 of Laspevresia were 
parasitized by one or the other of the parasites. 
The spruce-seed midge. Perrisia strobi NVINN. 
Larvae of gall-midges were already in 1848 known to live in spruce-tree cones, 
and in 1853 WINNERTZ described a gall-midge, Cecidomyzra strobi, bred by KALTEN- 
BACH from spruce-tree cones. NITSCHE (1895) found the larvae of a gall-midge in 
spruce-seeds, but did not succeed in breeding them, being therefore unable to 
solve the problem whether they were identical with Cecidomyvia strobi. No 
further attempts have been made in Germany to breed the spruce-seed gall-midge; 
and as late as in 1913 NUSSLIN quotes the statements of JUDEICH-NITSCHE, 
apparently not knowing that in 1890 J. SAHLBERG had sueceeded in breeding 
a gall-midge from spruce-seed, which he identified with Cecidomvia strobi. 
The knowledge of the spruce-seed gall-midges in Sweden has aiready been 
related. The opinion set forth by SYLVÉN that they were identical with /Ple- 
melhella abietina SEITN., was proved by the present writer to be false in his 
book The Forest Insects of Sweden”, but the species remained still un- 
identified. 
A detailed comparison between the spruce-tree gall-midge of Sweden and 
WINNERTZ' description of Cecidomyia strobi shows that both agree entirely in 
all essential details; and my identification has kindly been confirmed by J. 
KIEFFER in Bitsch, Lorraine, Germany. 
For details of the adult, the larva and the pupa, the reader is referred 
to figs. 15—20. 
The pupation. The fact that the investigators on one side found larvae 
in the seeds, which they failed to breed, on the other found cocoons in the 
scales from which gall-midges emerged, gave birth to the view that two dif- 
ferent species occurred in the cones. This, however, is not the case; the 
larvae feed on the seeds, leaving these before the pupation and entering the 
scales, where they form the characteristic white cocoons. 
The fact that LamPa found the larvae still unaltered in the seeds and that 
no one has managed to breed the larvae in Germany might be interpreted 
as evidence of one generation requiring more than one year. But, in my 
opinion, this feature is more probably due to the seeds having been kept 
too dry, the development of the larvae having consequently been arrested. When 
the seeds are subjected to germination experiments and kept moist the larvae 
leave them, evidently in search of some suitable place for the pupation. 
When the cones were kept in the cases described above they also got dry, 
it is true; but on account of the moisture they retained when enclosed in 
the cases this process went on slowly enough not to interfer vith the deve- 
lopment of the midges — hence the difference between the attempts to breed 
the midges when the cones are shelled or not. 
On examining a cone from which many gall-midges have emerged, no pu- 
pal exuviåe are found in the seeds, these remaining in the mouth of the bur- 
