310 Royal Society. 



In fig. 6, the central chord, just before its division into two 

 branches, throws off on each side a small branchlet ; in fig. 7 this 

 happens only on one side. Finally, figs. 4 and 7 present us with 

 some instances in which more than four branches are given off close 

 to the first division of the great chord C. 



But even in the case which I have above described as most 

 typical, the symmetry is not in fact so great as it w^ould at first 

 sight appear to be, because the nerves on the two sides are fre- 

 quently not of the same size. Thus, in fig. 2 each of the two 

 branches of the maui central stem divides, it is true, into two 

 secondary branches, one of which is smaller than the other, but the 

 two lesser branches are both upon the right side. If then, as is 

 probable, we are justified in concluding that in each animal the 

 ultimate nervous fibrils are of somewhat equal size, that they com- 

 pose the greater part of the nerve, and that the corresponding 

 organs of the two sides of the body receive an equal amount of 

 nerves, it is evident that some of the parts which on the left side 

 are supplied by the large outer branch must on the right side be 

 connected with the median branch. 



We see, therefore, that not only is the branching of the nerves 

 absolutely irregular, and that of the two sides entirely unsym- 

 metrical, but even the number of main stems proceeding from the 

 ganglion is not always the same. This result has surprised me very 

 much, since if any organs might have been expected to be almost 

 invariable, I should have thought it would have been the nervous 

 system. I believe that no parallel case has been described, nor do 

 I even remember to have seen a description of any variation occur- 

 ring in the larger nerves of any animal whatsoever. Considering, 

 however, how great are the variations which occur here in the same 

 species, it is evident that differences in the distribution of the nerves 

 in nearly allied forms are in themselves no proof ilidX such species 

 were separately created. 



Around the ganglionic masses are several large spheiical bodies. 

 These appear to be homologous with the " Zellenkorper," described 

 by Leuckart as surrounding the supracesophageal ganglion in the 

 larva of Melophagus. He considers them also as homologous with 

 similar organs which have been observed in the embryos of other 

 insects by Heroldt and Kolliker*. 



Dujardin (Ann. des Sci. Nat. 1850, 3 ser. vol. xiv. p. 202) de- 

 scribes the supracesophageal ganglion of the worker- ants as consisting 

 of several isolated parts, and I was at first inclined to consider these 

 spherical bodies as also merely isolated parts of the ganglionic mass, 

 in favour of which view it may be urged that fewer nerves than 

 usual appear to proceed from this mass. The contents of the 

 spherical bodies, however, under the influence of reagents, present 

 an appearance different from that of the supracesophageal mass. 



The suboesophageal ganglion is very richly supj)lied with tracheae, 

 derived from two large stems which are attached to the front angles, 

 and ramify from thence all over the mass. 



* Die Fortpflanzung iind Entwickelung der Piipiparer. Halle, 1858. 



