had to be added to. It was then found that the university had a much more valuable collection than was at first supposed. Among its most interesting and instructive features were a typical lot of about forty neolithic implements from Denmark; stone and pottery specimens illustrating the archæology of Michigan; potsherds and pottery vessels from the islands at the mouth of the Amazon; ancient Peruvian vessels from Ancon and Pacasmayo; pottery vessels from the East Indies; skulls, including some of the perforated crania described by Henry Gillman, so mounted that they may be examined on all sides and measured without handling; stone hammers or mauls from the ancient copper pits at Isle Royale, Mich.; the De Pue collection of implements from the surface of the immediate neighborhood of the university; and casts from the Smithsonian collections, with descriptions. Several prehistoric village sites have been discovered near Ann Arbor, which, it is hoped, may soon be thoroughly examined. A survey has been made of one of the prehistoric "garden beds" at Kalamazoo. As yet the work of the museum has been chiefly confined to archæology, but it is expected that, as the department develops, other phases of anthropological science will be studied.

The Conditions of Rain-making.—The conditions of successful rain-making, as defined by Prof. A. Macfarlane, of the University of Texas, in the light of Mr. John Aitken's experiments on fogs and dust—not in a small portion of the atmosphere cut off from the rest by means of an air-tight receiver, but on a large scale in the unbounded atmosphere—are: "If the air operated on is at a temperature higher than its temperature of saturation, it must be cooled down to that temperature. Further, when the moisture condenses it gives out latent heat, which tends to arrest the process; this latent heat must be removed. It is not, as some rain-makers have imagined, 'Pull a trigger; Nature will do the rest.' The only trigger-pulling which experiments warrant as possible consists in supplying the necessary fine dust for nuclei, so that condensation may take place without delay when the air is cooled to its temperature of saturation; or in supplying fine dust from such a substance as common salt, which has a chemical affinity for water and may be able to accelerate slightly the falling of a shower. Suppose we take a cubic mile of the air upon which Dyrenfurth operated on the night of Friday, November 25, 1892. The record at the weather office in San Antonio, at 8 p.m., gave the temperature of the air as 72° Fahr., and the dew point as 61° Fahr. To cool down a cubic mile of that air to the dew point would require the abstraction of as much heat as would raise eighty thousand tons of water from the freezing point to the boiling point. To cool it down another eleven degrees would require as much more heat to be abstracted. The amount of water set free would be twenty thousand tons, which, spread over a square mile, would give about 1·4 pound per square foot, or 27/1000 of an inch of rainfall. The amount of latent heat set free by the condensation of that amount of water would raise one hundred thousand tons of water from the freezing point to the boiling point; and it would be necessary to absorb this heat in order that the rain-making might go on. I have supposed the cubic mile of air to be kept constant; if the air operated on is constantly changing, the task becomes one of infinitely greater difficulty." It is hardly necessary to say that Prof. Macfarlane considers the professional rain-makers, the proceedings and pretensions of eight of whom he reviews in his paper, as "no better than the medicine men of the Indians."

Anomalies in Weight.—The anomalies in weight at different points on the surface of the earth, which have been recognized for a considerable time, have been attributed to corresponding anomalies in the figure of the earth; to the insufficiency of the formulas for reduction to sea-level; to the unequal distribution of masses; or to inexact observation. A study of the subject by the French commander Defforges, which included forty-one observations at twenty-five stations of different latitudes and elevations, shows that weight is distributed very unequally over the globe; that Clairaut's law, true as a whole, is nearly always marked by notable anomalies; that weight, on the shores of different seas, presents feeble anomalies, constant, and consequently characteristic, on