i88i.] 
( ii 7 ) 
NOTES. 
M. E. Brandt has laid before the Academy of Sciences an 
account of his very elaborate researches on the nervous system 
of the different orders of insects. He has examined 235 species 
of Coleoptera in their perfect state, and 36 in the state of larvae. 
He finds that in some ( Rhizotrogus solstitialis) the suboesophagian 
ganglion is blended with the thoracic. The cerebral ganglia have 
always convolutions. There are from one to three thoracic 
ganglia. The abdominal ganglia vary from one to eight, the 
number varying sometimes in the two sexes of the same species. 
Sometimes, as in the Curculionides and Lamellicornes, there are 
no distinct abdominal ganglia, but they are blended with the 
thoracic portion. The author's results on the nervous system of 
the Hymenoptera were published in the “ Comptes Rendus ” 
(lxxxiii., p. 612). All the Lepidoptera have two cephalic ganglia, 
the upper one having convolutions. There are always four ab- 
dominal ganglia, save in Hepialus humuli, where there are five. 
The Diptera have always two distindt cephalic ganglia, the upper 
being convoluted. The thoracic ganglia vary from one to three, 
and the abdominal from one to eight, the number sometimes 
varying according to sex. Sometimes also the abdominal and 
thoracic ganglia are blended. In the Hemiptera the sub- 
cesophagian ganglion is sometimes placed in the thorax : they 
have never distindt abdominal ganglia. 
MM. Garreau and Machelart have extradted from plants of the 
Saxifrage group an alkaloid which they name benzonine, and 
which in its physiological adtion has some resemblance to 
quinine. 
Mr. S. H. Scudder has produced a memoir on the fossil insedts 
from the Devonian of New Brunswick. We make the following 
extracts from the report given by the “ American Naturalist — 
The general type of wing-strudture has remained unaltered from 
the earliest times. Three of these insedts ( Gerephemera , Homo- 
thetus, and Xenoneura) have a neuration differing both from the 
carboniferous and the modern types. The earliest insedts were all 
hexapod, all hetero-metabolous, and all allied to the Neuroptera. 
They show marks of affinity to the carboniferous Palceodidtyoptera, 
but are often more complex in strudture. They differ remarkably 
from all other known types, ancient or modern, and some of them 
appear more complicated than their nearest living allies. While 
there are some forms which bear out expectations based on the 
general derivative hypothesis of structural development, there 
