276 CLASS VIII. 
The reproductive power in Insects which undergo metamorphosis 
is wanting in their perfect state; but if at an earlier period in 
the state of larva they have lost a foot, it grows out again at the 
next moulting, and is more or less perfectly restored. Also in the 
Myriapoda excised antenne grow again'. 
The nervous system of Insects has for central part a row of 
ganglia of different number, which are usually connected with one 
another by two threads that are often very intimately united. This 
row of ganglia is situated on the ventral surface beneath the imtes- 
tinal canal in the mid region of the body ; the first ganglion however 
lies in front of and above the cesophagus, and there arises, by reason 
of the two threads which connect it with the second ganglion, a 
ring which surrounds the cesophagus. The greatest number of 
ganglia is found in the Myriapoda, eighteen in Lithobius (Scolopendra 
forficata), twenty-three in Scolopendra morsitans. In the larvee of 
butterflies thirteen are counted, but ordinarily they are less numerous 
in the hexapod Insects. Large ganglia are situated in the thorax, 
and in some there are none in the abdomen, but two nervous strings 
alone, sometimes close together, sometimes separate from each other, 
as in Nepa and Cicada. From the ganglion above the cesophagus 
(ganglion cerebrale) arise the nerves of the eyes and antenne; this 
ganglion lies transversely on the cesophagus, formed of two oval or 
somewhat conical lateral portions with their broad part turned to 
each other; the inferior surface is somewhat concave, the upper 
convex. The second ganglion, the first of those beneath the intestinal 
canal, is by some writers compared to the cerebellum, by others, on 
better grounds, with the medulla oblongata; the nerves that arise 
from it proceed to the oral parts, and perhaps correspond to the 
different branches of the fifth pair in vertebrate animals. Harlier 
writers, as ACKERMANN, Rum and BicuAr, thought that the abdo- 
minal cord of Insects might be compared with the nervus sympa- 
thicus of vertebrate animals; CUVvIER and GALL, on the contrary, 
have disowned and rejected this correspondence. It is necessary in 
this inquiry to determine in the first place what character is to be 
considered of sufficient value to distinguish the spinal cord from the 
1 Newport made experiments on Julus, Lithobius and caterpillars of butterflies. 
See Phil. Trans. 1844, p. 283. In Phasma sometimes one foot is less than the rest, 
being a new growth. I found this once also in Reduvius personatus. 
