Miscellaneous. 505 



In each form of the nervous system there is a different number of 

 abdominal ganglia (thi'ce to seven), while the larvse have eight 

 abdominal ganglia (except the larva) of the Pteromalidae, which 

 have no chain of ganglia, but a simple and compact nervous mass 

 as in the larvae of the flies). During the pupa stage the number of 

 ganglia diminishes in the diff"erent species, many of them approaching 

 one another and becoming fused. The supraoesophageal ganglia are 

 very strongly developed and completely cover the small subcesophageal 

 ganglion, which is united to them by very short commissures. The 

 examination of the pedunculate bodies has shown me a very singu- 

 lar peculiarity which has not previously been observed, F. Dujardin 

 remarked that the development of these bodies corresponds with 

 the degree of development of the instincts and intelligence in the 

 different species ; my researches prove that this is also the case for 

 the different sexes of the same species. Thus in the workers of 

 the common bee they are of immense size, while they are slightly 

 developed in the queen and in the males ; this is the case also in 

 the wasps and the ants. The pedunculated bodies do not emit 

 ocellar nerves as P. Dujardin has asserted ; these latter emerge 

 from the upper part of the supraoesophageal ganglia. 



The subcesophageal ganglion is very small, formed of a pair of 

 nuclei, and gives origin to three pairs of buccal nerves. Where the 

 nervous system has three thoracic ganglia, the first and the second 

 are simple, and have only two nuclei, while the third is always 

 more or less composite. In the Phytospheces there are two pairs 

 of nuclei, and in the Entomospheces, as well as in Cerceris, Pompilus, 

 Ammoj)hila, and Formica, three pairs. It is evident that in the 

 first case the last thoracic ganglion results from a fusion of two, 

 and in the latter case of three ganglia of the larva. In the Hyme- 

 noptera which have only two thoracic ganglia (bees and wasps), 

 the second presents four pairs of nuclei, resulting from a fusion of 

 four ganglia of the larva (the last two thoracic and the first two 

 abdominal ganglia). The number of abdominal ganglia varies from 

 three to seven. Hitherto it has been thought that only the last 

 abdominal ganglion is composite, while the others are simple ; but I 

 demonstrate that, in many cases, it is the penultimate abdominal 

 ganglion which is composite (the worker bee, the female of Mutilla 

 europoiu), while the last is simple. The largest number of abdo- 

 minal ganglia (that is to say, seven) exists in the lower representa- 

 tives of the order Hymenoptera, the Phytospheces, in which all 

 these ganglia are simple as in the larva?. Most of the Entomo- 

 spheces (Ammophila, Cerceris, Odynerus, and Bomhus) have six 

 simple abdominal ganglia. If there are only five abdominal ganglia, 

 two different forms are found : it is either the last abdominal 

 ganglion (Andrena and the worker of the wasp) or the penultimate 

 ganglion which is composite (the worker of the bee). Where there 

 are only four abdominal ganglia, it is usually the last that is com- 

 posite. In the Encerata and Crabrones {Ectennins, Thyreopus, &c.), 

 having only three abdominal ganglia, the last, which is always very 

 large, is produced by the fusion of the last four ganglia of the larva. 



Ann. (& Mag. N. Hist. Ser. 4. Vol. xviii. 34 



