ANATOMY or SPHiEEOTHERIUM. 169 



large Spiroholus from Ceylon I found four ; and in different 

 species of Jiduft only two or three. These projections are little 

 conical caps of chitin perforated at their extremities, and contain- 

 ing the terminal nerve-fibres of tlie organ of sens^e. It was very 

 difficult to obtain good sections of this organ. The only method 

 was to imbed the antenna in paraffin and then to cut close round 

 it and pick away the chitin bit by bit, afterwards re-embedding 

 the soft parts, and cutting a series of sections in the ordinary way. 

 The antennae of my specimens being small as compared with the 

 bulk of the body, were better preserved and fit for histological 

 purposes. 



The antennary nerve breaks up in the penultimate joint into 

 as many branches as there are nerve-endings on the terminal 

 joint, and each branch is continued into a spindle-shaped bundle 

 of nervous tissue which lies in the terminal joint. Thus there 

 are a number of these bundles lying side by side in the terminal 

 joint of the antenna, and each mass lies directly underneath, and 

 is continued into one of the conical projections described above. 



Each bundle of nerve-fibres, as it passes into a spindle-shaped 

 body, breaks up into a number of branching and anastomosing 

 nerve-fibrils, forming a neurosponginm. from the neurospon- 

 gium proceeds a bundle of nerve-fibres, which are beset in their 

 course with a number of ganglion-cells. These cells are nucleated, 

 and lie along the nerve-fibres so as to give the appearance of re- 

 gularly arranged parallel linear series of cells (see figs. 11 and ] 2). 

 The nervous bundles are thickest in the region of these nerve- 

 cells ; beyond them the bundle tapers away somewhat rapidly, 

 and consists of closely-packed fine nerve-fibres which run up 

 towards the conical projections on the end of the antenna, and 

 enter into close connexion with the bases of the fine sense-hairs, 

 which protrude slightly through these projections. The bundles 

 are isolated from one another by connective tissue, and fine 

 tracheal branches run between them riglit up to the end of the 

 antenna. 



Lying among the branched connective tissue-cells and sur- 

 rounding the proximal ends of the nervous bundles are a number 

 of large oval cells, each with a distinct oval nucleus, which stains 

 deeply with hsematoxylin or borax-carmine. Whatever the 

 nature of these cells may be, they are certainly not ganglion- 

 cells, and give off no processes continuous with the nerve-fibres 

 (%. 12, 0- 



