ZOOLOGY AND BOTANY, MICROSCOPY, ETC. 169 



SchUbeodes found in North American streams. These catnshes have an 

 axillary pore, which is the opening of a gland. Experiments with 

 S. gyrinus indicate that the secretion of the gland is poisonous. In 

 addition to the axillary glands, S. gyrinus and S. noctunrus have similar 

 glands developed about the pectoral and dorsal spines, with ends pro- 

 jecting slightly through a slit in the epidermis. Spine-glands are not 

 found in those species which possess well developed seme upon the 

 spines. 



The glands are invaginations of the epidermis ; the gland -sheath 

 is modified corium ; the clavate cells of the skin become the secretory 

 cells ; the ordinary epidermal cells form a supporting network ; there 

 are no muscles for forcing out the secretion ; the cell-walls are evidently 

 ruptured by the pressure of their contents, and in this way the spines are 

 constantly anointed with the poisonous secretion. 



Poison Apparatus of Weever.* — J. 0. Borley describes the poison- 

 glands of Trachinus draco and T. vipera, which are lodged in five or 

 six rays of the dorsal fin and in a spinous outgrowth of the opercular 

 bone. The opercular gland consists of a capsule of connective-tissue, 

 a rich network of capillaries, and very large secretory cells in radiating 

 columns. The secretion appears in two states : masses of finely granular 

 material, and highly refringent colloidal substance, either two secretions 

 or two stages of one secretion. It is highly probable that there is a 

 perpetual waste of secretion into the sea, though this is minimised by 

 the closeness with which the sheath fits the spine. Where the spine 

 issues from the substance of the operculum it is still at the bottom of 

 a tube sunk in the operculum, this tube being the sheath. This tube 

 wrinkles down about the spine as the latter enters a victim until about 

 one-third of the spine is uncovered. 



H. Muir Evans f has made some experiments on the action of the 

 weever's poison. He refers to the previous investigations of BottardJ 

 and Briot,§ but his own work was independent of these. An injection 

 of the poison into gold-fish, frog, mouse, and guinea-pig, produced local 

 paralysis. Marked haemolysis was seen in the blood of pigeons and 

 various mammals. The poison is probably an "amboceptor," which 

 unites with the endocomplements of the blood-cells. 



Food of Birds. || — Cecil H. Hooper has gathered together a number 

 of facts in regard to the food of birds, especially of those that are 

 important practically. A few examples may be given. The amount of 

 insect-food eaten by sparrows is comparatively small. Bullfinches do 

 much harm to fruit-buds, especially gooseberries. Blackbirds destroy 

 much fruit, but are harmless or useful at other times, eating worms, 

 grubs, etc. Starlings devour leather- jackets and wireworms, but destroy 

 much fruit. Missel-thrushes eat many fruits, but outside the fruit 

 season they do no harm. The song-thrush devours fruits, but also 

 insects, snails, and worms. Greenfinches are a terrible pest among hops ; 



* Trans. Norfolk and Norwich Nat. Soc, viii. (1907) pp. 369-73 (1 fig.). 

 t Tom. cit., pp. 355-68 (1 fig.). $ Les Poissons Venirneux, 1889. 



§ C.R. Soc. Biol. Paris, 1902-4. 

 || Journ. Board of Agriculture, xiv. (1907) pp. 402-12. 



April loth, 1908 n 



