other bear is largely a fruit, vegetable, and ] insect feeder; but in the frozen north the polar bear lives by necessity mainly on fish, carrion, seals, walruses, and birds. Its notion of an 'egg for breakfast' is rather amusing. It will clear an islet of eider-ducks' eggs in a few hours."
Dispersal of Fresh-water Shells.—A recent book by Mr. H. W. Kew deals with the means by which fresh water and land shells are dispersed. The occurrence of these shells is sometimes very puzzling, as when fresh-water shells are found in isolated ponds. It is surprising how varied the means of distribution are. The animals are carried down stream on various floating objects. A case is cited by the author in which a number of anodons were carried away by a whirlwind and fell with the rain. Canon Tristram found the eggs of a mollusk attached to the foot of a passing mallard which he shot in the Sahara a hundred miles from water. A few instances have been noted in which birds on the wing have been shot with bivalves adhering to their toes. Insects also lend their aid, and a water-beetle has twice been captured on the wing with Sphœrium attached to its legs. Another specimen was caught with Ancylus attached to its wing-case, and other aquatic insects have been found with mollsks attached to them. The actual process of transportation of land shells has not often been observed. Some live snails were once found in the stomach of a wild pigeon three days after it had been shot, and an operculated land snail has been found dragged along on the foot of a bumblebee on which it had caught. An isolated dew pond after an existence of ten years will generally yield several species of fresh-water mollusca, and a mediæval fish pond has a considerable fauna. A church or castle built of limestone, but surrounded by non-calcareous desert, is for a large group of land snails the equivalent of an isolated pond; but it is only on very old buildings that one finds colonies of the special limestone species. Mr. Kew also discusses the dispersal of shells by human agency.
Hearing of Infants.—In her Notes on the Development of a Child, issued in the series of University of California Studies, Milicent W. Shinn reports that the infant started violently while nursing, when a paper was torn some eight feet away, on the third or fourth day after birth, and at several times on that and the few following days she started and cried out even in sleep when a paper was rustled sharply as her father sat by the bed. During the first week she did not seem to notice when on his return in the afternoon her father sat close by, reading aloud or talking, but in the second and third weeks she always became restless at this time. The more modulated voices of women who were in the room the rest of the time appeared not to affect her at all. The sensitiveness to sound seemed variable, for on the twenty-third day, when Miss Shinn purposely rustled paper near the baby, it produced no clear reaction, nor did a table call bell struck suddenly and sharply at two feet and even one foot from her head. On the twenty-seventh day she showed no sign of hearing single notes on the piano from the highest to the lowest, yet she started at a hand-clap behind her head. Ten days later, while the baby was lying half asleep on Miss Shinn's lap, the servant brought in a tin bath tub and set it down abruptly so that the handles rattled. The infant started violently with a cry so loud that it brought in her grandfather from two rooms away to see what was wrong. She also put up her lip with the first crying grimace she had ever made, and showed the effect of the fright in a disturbed face for five minutes. Yet throughout the first two months there were also many times when she failed to pay any attention to sounds quite as striking as the few she did notice. The great variation in sensibility was especially noticeable in the second month.
The Tropical Climate.—Respecting the climatology of tropical Africa, E. G. Ravenstein represents that by ascending a mountain we might, even in tropical Africa, enter a region the mean temperature of which coincided with that of England; but if we at the same time considered the annual and daily ranges of temperature, we should find that a tropical climate differed exceedingly from that of the temperate regions. In the latter the annual range was considerable, the daily range small. The character of a tropical climate was the very reverse, for there
