In order to improve on these charts, he proposes to extend his daily weather-charts by extra- and interpolation over those portions of the ocean from which he has no direct observations, and to combine into monthly means the readings taken from his daily charts. At present he has charts for only two years-1874 and 1875-which he gives, and which, of course, show a distribution of pressure decidedly different from that given in his general chart. The charts are also given for December, 1874, and February, 1875; that for December showing a very remarkable departure from what would be considered a normal distribution.

He seems to consider that three minima—namely, that of Davis Strait, that northeast of Iceland, and that southwest of Iceland-by varying their position and relative development, alternate in their control over the winds and weather of the Atlantic.

The character of the weather which prevails over Northern Europe depends, therefore, entirely upon the predominance of one or other of the barometric minima of the North Atlantic Ocean. While recognizing the fact that, theoretically, the wind affects the distribution of pressure, he seems to only partially appreciate the importance of the law established by theoretical mechanics—namely, that it is not so much the barometric pressure that determines the wind, as it is the wind that determines the pressure.

He recognizes the fact that the area studied by him is but a very small portion of the northern hemisphere, and that the causes of the important variations in the distribution of pressure must be, at least in part, looked for outside of the region covered by his maps. In this respect, certainly, the monthly maps published by Ferrel in 1877, and the daily weather-maps of the northern hemisphere published by the Signal-Office in 1878, must be recognized as the first steps in the proper treatment of this subject.

EVAPORATION AND PRECIPITATION.

Weilenmann, in the Schweizerischen Met. Beob., 1877, vol. xii., gives the development of a formula for the quantity of evaporation, which agrees, in the most remarkable manner, with observations at Vienna, St. Petersburg, Mont Souris, Pola, and Tiflis. This memoir and his essay on Atmospheric

Temperature constitute two important contributions to deductive meteorology.

Professor Nipher sends us, among the publications of the Missouri Weather Service, a valuable table of monthly, annual, and seasonal amounts of rainfall observed at St. Louis, principally by Dr. George Engelmann, from 1834 to 1877. There are but few stations in the world that can present an unbroken homogeneous series like this, and it is to be hoped that similar tables may be published for such other long series of observations as we may have in the United States. Such a collection, supplementary to the Smithsonian Tables, would be useful in many investigations.

A paper, by Otto Krümmer, on the Distribution of Rainfall in Europe, is published in the July number of the Journal of the Berlin Gesellschaft für Erdkunde.

Dr. G. Hellmann has published, in the Netherlands meteorological Jaarbooek, a work on the Humidity and Cloudiness of Spain and Portugal. He gives the hourly variations for 4 stations, and the monthly and annual means for 18 stations, for 12 years. The average number of cloudless days during the year is, for Obiedo, 50; Saragossa, 199; and Valencia, 260.

The distribution of rain over Germany, according to the four seasons, is almost exhaustively treated of by Dr. J. Van Bebber, in Petermann's Mittheilungen, and is accompanied by four charts, showing isohyetals for each 25 millimeters. He distinguishes, for Germany, three well-marked regions: first, the west coast, or region of heavy autumnal rains; second, Alsace, the region of heavy winter rains; third, the region of summer rains, which includes pretty much all the rest of Germany.

An investigation, by Dogiel, published in Vol. XX. of the Bulletin of the St. Petersburg Academy, into the innumerable forms of the hexagonal crystals of iodoform (CHI), affords additional reasons for careful investigation into the circumstances that determine the formation of the varieties of snow-crystals. It is highly probable that definite temperatures and pressures may be indicated by these forms.

The large number of observers of rainfall in India will surprise every one who has not especially looked into this matter. The government supplies rain-gauges to all districts and

subdivisions. They are also generally placed at every planter's station, and at the tea-gardens, under the control of the various tea companies. Several provinces have more than 100 gauges each, and in all India there are estimated to be at least 1500 rainfall stations.

This subject was brought to notice by the reading of a paper, by C. N. Pearson, on the Meteorology of Mozufferpoor, Tirhoot, India (Quarterly Journal of the Meteorological Society, 1877, p. 410). An extraordinary rainfall of 14 inches in 12 hours occurred between midnight of the 22d of September, 1876, and the following noon, and measures at seven neighboring places showed the local nature of the rain.

Returns from 900 rainfall stations in France are published quarterly by the Association des Sciences de France and the Département de Ponts et Chausses under Belgrand.

Mr. Nathan Butler communicates to the Bulletin of the Minnesota Academy some notes on a hail-storm which he experienced in the western part of Minnesota on the 18th of August, 1858. The sky was generally clear, the weather quite warm, and the clouds overhead very light and fleecy. Immediately following a flash of lightning, large hailstones began to fall, and continued for perhaps two or three minutes. They buried themselves for about half of their diameter into the sod of the prairie. When the shower was finished, the stones were sprinkled into the ground about fifteen feet apart, and the larger ones were about the size of a man's two fists. In shape they were spherical on one end, made up of hexag onal crystals, like crystals of quartz; the other end was conical, made up of white ice. They were quite solid, and did not break in falling. They were found to weigh a pound each.

The abnormal character of the weather of the winter of 1877-78 is strikingly seen in the immense floods of the Sacramento valley. It is probable that such floods may occur every century. The enormous erosions west of the Rocky Mountains may be due to such occasional floods quite as much as to any regular annual rainfall.

A very complete synopsis of the various theories with regard to the formation of hail is given by Dr. T. H. Bauermeister in successive numbers of Dr. Klein's excellent popular scientific journal Gaea. He reviews the history of the subject from the time of Musschenbroek, and gives especial at

tention to the views of Volta, Leopold Von Buch, Muncke, Mohr, Reye, Lucas, Baumgartner, and Dellmann. He concludes, however, by saying that previous to any critical discussion of these different theories, it is, above all, necessary that there should be as full as possible a collection and short description of all hail-storms that have been scientifically observed with reference to the form of the clouds, the direction of the wind, the atmospheric electricity, the topographical distribution, as well as a collection of the results of scientific balloon voyages, and as many as possible observations on high mountains. He hopes in a short time to collect such material, and will gladly welcome any contribution.

Reynolds has considered the formation of hail-storms, and has artificially reproduced them in great perfection.

STORMS.

Professor Loomis has published his eighth contribution to meteorology, January, 1878, and his ninth paper in July, while the tenth paper is understood to be in press, having been read before the National Academy in October, 1878. The eighth paper deals with the origin of areas of low pressure, and the ninth paper considers especially those storms that come from the Pacific coast eastward to the Mississippi valley. He then takes up the areas of high barometer. Combining the two studies with the observations of the clouds, he arrives at a general circulation of the atmosphere similar to the views published by Ley, Hildebrandsson, etc.

The important memoir of Captain Henry Toynbee on the Meteorology of the North Atlantic during August, 1873, has been published by the Meteorological Council. This work relates especially, of course, to the history of the great hurricane of that month, and does not appear to have been prompted by the report of Mr. Abbe on the Nova Scotia hurricane, but to have been begun in December, 1873, quite independent of, and before the publication of, that work. Captain Toynbee has collected some 280 ships' logs, and has utilized all other sources of information; so that his work is undoubtedly the most valuable collection of facts relating to any hurricane which has ever yet been published, and is really a monument to his patience in collating material which, judging from our own experience, must have offered very many

perplexing discrepancies. His work includes 31 daily charts of the North Atlantic and a concluding synoptic-chart, while his text presents us with a review of each daily chart in succession, followed by a series of chapters on the hurricane, and conclusions that can be drawn from the study.

He states that from the 1st to the 10th, the northeast trade and southwest monsoon were often in close proximity over that part of the sea which lies to the southwest of Cape Verde. It seems most probable that the hurricane was forming on the 12th in about 11° N. and 20° W. From that time on, the track of the centre of the hurricane is occasionally indicated until the 17th, from which date until the 25th its track is very clear as it passed westward between the West Indies and the Bermudas, and then turned northward, and finally northeastward, near the southern coast of Nova Scotia. On the 26th he concludes that the centre of the hurricane was south of Newfoundland, but that it was broken in force, and that contact with land broke up the great eddy, which had shown no signs of breaking up so long as it was over the open sea. From the 27th to the 31st there appear areas of slight barometric depression between Labrador and Europe, attended occasionally by strong gales but no hurricane. By means of these depressions, which represent the breaking-up of the hurricane, the meteorologist would continue the track of the hurricane eastward to Great Britain; but the practical seaman would say that the hurricane, as such, died out in Newfoundland. He shows that had we telegraphic communication with the Bermudas and St. Thomas, timely warnings might have been sent to Nova Scotia and the United States. The law according to which areas of low pressure in Europe pass outside of, instead of advancing into, areas of high pressure seems to have also prevailed in the progress of this hurricane.

The important question with regard to the direction of the wind and the bearing of the centre of the hurricane is elucidated by diagrams, and especially by measures made on three different charts, and at distances from 100 to 800 miles from the centre. On these charts the angle between the bearing of the hurricane-centre and the direction of the wind has been carefully measured for 108 different ships' positions, the average of all of which shows that the wind does not