1879.] NATURAL SCIENCES OF PHILADELPHIA. 1G1 



above referred to showed that it had been penetrated, as far as ex- 

 amined, by the water, as the soil was thoroughly soaked through. 

 The numerous galleries must give more ready access to the rains 

 within formicary bounds than elsewhere. The points are of 

 much interest, and are still under examination, but the following 

 suggestions were made : First, the peculiar arrangement of the 

 galleries and chambers indicates that the least exposed portion of 

 the nest is that near the surface, in the parts which do not com- 

 municate directly with the same. The main entrance and exit 

 being removed from this, and penetrating downward and heneath 

 it, would cause a drainage which, carrying oft" the first flood, would 

 leave the upper chambers comparatively safe until the water should 

 fill up the lower spaces and back up to the surface rooms. Second, 

 it is probable that the galleries which penetrate downwards may 

 serve the purpose of drainage downward. In heavy rains, how- 

 ever, neither of the above arrangements woidd seem to afford am- 

 ple protection. Third, it is therefore probable, and observation 

 and some experiments 1 appear to point in this direction, that the 

 ants themselves (if not the larva?) can endure a submersion more 

 or less prolonged with comparative safety. 



Honey Glands on Catalpa Leaves. Mr. J. A. Ryder stated at 

 the meeting of the Botanical Section that he had recently observed 

 the presence of a number of large nectar-secreting glands on the 

 under side of the leaves of the common Catalpa bignonioides. 

 These glands are situated in the axils of the veins of the leaf, i. e., 

 where the lateral veins join the midrib. Those nearest the inser- 

 tion of the petiole are largest, whilst toward the apex of the leaf 

 they are smaller. The glandular areas, extending over a consider- 

 able axillary space as well as to some extent over the sides of the 

 veins, are without hairs, the place of the latter being taken by 

 large biscuit-shaped, sub-circular glandular bodies attached to the 

 surface of the leaf much like a button to a piece of cloth, and 

 projecting above the circumjacent epidermis, though at the point 

 where the gland is attached the epidermis is depressed. The ap- 

 pearance is not much unlike that of the circumvallate papillae of 

 the base of the human tongue. The glands seemed to be modified 

 hairs, and in thin vertical sections are seen to be composed of 

 columnar cells arranged around a cavity. The nectar observed in 

 a few instances was perceptibly sweet to the taste, and thrown out 

 in sufficient quantity to be seen as small clear drops in the axils 

 of the veins. Ants of both red and black species were seen feed- 

 ing upon this sweet liquid with great avidity. 



1 See a brief paper on the Vital Power of Insects, Proc. Acad. Nat. Sci. 

 of Phila., 1877, p. 134. 



