604 MR. E. B. POULTON ON THE [DeC. 18, 



(even under this power) basement membrane. The fibrils often 

 expand at their termination, becoming funnel-shaped. Such appear- 

 ances are observed over the whole proximal surface of the bulb and 

 not merely at its centre. It was really impossible to be certain iu 

 the identification of these fibrils as nervous, and yet there are some 

 fibrils which can be considered nervous with a very high degree of 

 probability. These are distinct under comparatively low powers 

 as sharply defined dark fibrils that pass straight through this layer 

 from the fibrous layer towards the taste-bulbs. These fibrils confer 

 a radiate appearance upon the subepithelial layer (see fig. xxi. Plate 

 LV.). Eound or oval, deeply-staining nuclei are very character- 

 istic of this layer, and are obviously related to the nervous elements, 

 as they are almost completely limited to the region of the end- 

 organs. These nuclei belong to small multipolar cells continuous 

 with some strands of the network ; but it seems more probable that 

 they belong to a special supporting connective tissue (such as the 

 neuroglia) than that they are nervous. This subepithelial layer 

 bears a strong resemblance to certain retinal layers, and is probably 

 identical in structure (both consisting essentially of the unravelled 

 ele'ments of supporting and nervous tissues arranged in fine 

 interpenetrating networks, as has been mentioned). 



These layers and their relation are better seen in longitudinal 

 than in transverse sections of the papilla (see fig. xx. Plate LIV., 

 which shows the same arrangement in Phalangista). 



B. The Lateral Gustatory Organs. — When a section is taken at 

 right angles to the long axis of one of the depressions (upon one of the 

 elevations previously described), the latter is seen to be the mouth 

 of a narrow chink which is obviously the duct of a gland (see fig. 

 XXXI. Plate LV.), in the epithelial walls of which a few taste-bulbs 

 have been developed. The lateral ducts sometimes open into the 

 chink above the taste-bulbs, and below the points where the latter 

 occur the main duct breaks up into smaller tubes. Horizontal 

 sections show that the narrow ducts into which the depressions open 

 are always slit-like, although the latter may appear to be circular, 

 and the long axis of the slit is always at right angles to the inferior 

 limit of the papillate surface above. In this respect the primitive 

 lateral structures of Hahnaturus are similar to the furrows of the 

 lateral organ in higher animals. Although the bulbs are scattered 

 irregularly in vertical sections, their arrangement is much more even 

 in sections taken horizontally. It is therefore probable that the 

 real arrangement is in regular tiers, but that the tiers themselves do 

 not follow one another regularly. The subepithelial layer is not 

 strongly developed beneath these bulbs, but traces of it can be 

 distinguished. The nerves approach the bulbs from the sides, 

 running horizontally beneath the epitheUum for a considerable 

 distance. Granglion-cells are very numerous in little groups on the 

 nerves. The cells are enclosed in distinct nucleated capsules. The 

 glands into which the ducts of the lateral organs lead are of course 

 serous. The smaller gland-ducts which open below and in front of 

 the lateral organs (see fig. i. &c. Plate LIV.) lead down deeply into 



