62 



NATURE 



[May 17, 1900 



Experiments on the exposure and development of photo- 

 graphic plates in ordinary light have recently been made by 

 Prof. F. E. Nipher, and are described in Science. It appears 

 that if a photographic plate in a camera is greatly over-exposed 

 it may be developed in the light. A plate which should for 

 ordinary work have an exposure of a second and a half for 

 street or outdoor photography, may be exposed for two hours. 

 When developed with weak hydrokinone by the light of a lamp, 

 it gives a good positive. If the plate is held too near the lamp 

 the light will dissolve a picture already appearing. If held too 

 far away the plate begins to fog. By moving toward or from 

 the lamp the proper illumination may be soon secured. It is 

 remarkable that a street scene taken in this way shows not a 

 moving thing on the streets. In Prof. Nipher's pictures, tram- 

 cars passing every two minutes, waggons, horses, pedestrians, 

 left no trace upon the plate. But the fixed objects are shown 

 perfectly, with their proper shadows and high lights. Prof. 

 Nipher points out that lantern slides and transparencies may 

 be made directly by this method without re-photographing from 

 a negative. Rontgen ray pictures can also be obtained upon 

 plates which have been exposed to the light of an ordinary j 

 room for a few days, by developing in the manner described. 

 Good radiographs have been thus produced upon plates which 

 were uncovered during exposure to the rays. 



The usual proof that the arithmetic mean of any number of 

 positive quantities is greater than their geometric mean consists 

 in showing that if any two of the quantities be replaced by their 

 semi-sum, the new series has the same arithmetic mean and a 

 greater geometric mean. This proof, however, involves the 

 assumption that if this process of substitution be repeated in. 

 definitely, the ultimate result will be a series of quantities each 

 equal to the arithmetic mean of the original series. We have 

 never seen this property proved, and it is certainly by no means 

 an obvious truth in the general case, for the result of the repeated 

 operations is always a fraction whose denominator is a power of 

 2, while the arithmetic mean of n quantities has n for its de- 

 nominator. We are glad to see that Mr. G. E. Crawford, 

 writing in the Proceedings of the Edinburgh Mathematical 

 Society, recommends an alternative proof in which the number 

 of steps is finite, and the above assumption is not made. Two 

 such proofs are possible, both of which run on somewhat parallel 

 lines, and Mr. Crawford refers to a text-book which appeared a 

 few years ago for the alternative to the proof now given. 



We have received from Prof. A. Klossovsky, the energetic 

 director of the meteorolo;^ical system of South-west Russia, a 

 very valuable contribution to climatology. The work consists 

 of two volumes, text and charts, and embraces the large area 

 running from about the latitude of London to the northern 

 shores of the Black Sea and the Sea of Azov, and bounded on 

 the east by the River Dnieper. The observations used in the 

 discussion include those made at the stations belonging both to 

 the Central Meteorological Service of St. Petersburg and to the 

 South-west Russian system, and embrace a period of twenty-five 

 years (1871-1895). The tables exhibit monthly, yearly and 

 five-yearly values of all the principal elements, and the distribu- 

 tion of thunderstorms and hail. The charts are coloured, and 

 show very clearly the mean annual distribution of rainfall, the 

 number of days of thunderstorms and hail, mean and extreme 

 temperatures, and the distribution of cloud and humidity. The 

 tables are arranged in various ways, and furnish most useful 

 statistics for agriculturists and for men of science generally. 



In describing some Neocene corals of the United States {Proc. 

 U.S. National Museum, vol. xxii. 1900), Dr. H. S. Gane re- 

 marks that a majority of the corals in these Eocene, Miocene 



NO. 1594, VOL. 62] 



and Pliocene formations belong to extinct species. They do 

 not, however, present any close kinship with the corals of a like 

 age in the West Indies, but are more nearly related to tho e 

 now living in the Caribbean Sea and Atlantic Ocean. 



Mr. Cecil B. Crampton, who has for some time been 

 assistant to Prof. Boyd Dawkins in the museum at Owen's 

 College, Manchester, has been appointed an assistant geologist 

 on the Geological Survey of Scotland. 



Mr. Lester F. Ward gives an account of the wonderful 

 " Petrified Forest " or " Chalcedony Park " of Arizona (Report 

 toDepartmentpf the Interior, U.S. Geol. Survey, 1900). Count- 

 less logs of silicified wood occur over a wide area in Arizona, 

 but they are especially abundant in a particular tract known as 

 the " Petrified Forest," east of Holbrook, between the Little 

 Colorado and Rio Puerco. Here the logs lie in the greatest 

 profusion, " while the ground seems to be everywhere studded 

 with gems, consisting of broken fragments of all shapes and sizes, 

 and exhibiting all the colours of the rainbow." These silicified 

 blocks are not in sittt, but have been derived from a bed of 

 conglomeratic sandstone of Triassic age, which is exposed on 

 the margin of a high plateau. Mr. Ward refers also to a well- 

 known "Natural Bridge," which consists of a petrified trunk 

 lying across a canyon, and forming a footbridge, and he observes 

 that the trunk here is in sUit. He advocates that means be 

 taken to preserve these natural phenomena. 



A report on the proposed railway from the Commune de* 

 Houches, Bonneville, in Haute- Savoie, to the summit of Mont 

 Blanc, has been published by M. Joseph Vallot, Director of 

 the Mont Blanc Observatory, and M. Henri Vallot, engineer. 

 This great undertaking was projected by M. Saturnin Fabre, 

 but various routes have been suggested. These are fully dis- 

 cussed by the authors, who give reasons for recommending a 

 route which starts from the valley .of the Arve at an elevation 

 of about 3000 feet, and proceeds by the Aiguille du Gouter and 

 the Dome du Gouter to a terminus at the Petits Rochers 

 Rouges, where the elevation is about 15,000 feet. The total 

 length ot the railway would be about seven milesi and from an 

 elevation of about 4000 feet to its upward termination, the line 

 would for the most part be subterranean. There would be 

 several openings, and also stations giving access to the mountain, 

 at points of special interest and beauty. M. Joseph Vallot 

 contributes chapters on the geology, including the glacial 

 phenomena, and these are illustrated by a section showing the 

 nature of the solid rocks through which the railway would be 

 carried, and the thickness of the glacier-ice above. For a short 

 distance the tunnelling would be made through Liassic slates 

 and Trias with gypsum, and then wholly through various 

 crystalline schists. j 



In his memoir recently published in the Philosophical Trans- 

 actions (see Nature, April 19, p. 595). Mr- Oldham has shown 

 that, in recording the movements due to distant earthquakes, 

 the heavy vertical pendulums employed in Italy answer most 

 readily to the early tremors, while the light horizontal pendulums 

 of Rebeur-Paschwitz and others are most affected by the later- 

 arriving surface- undulations. Dr. G. Agamennone has discussed , 

 the same subject independently in a note read on February iS 

 before the R. Accademia dei Lincei of Rome. At the Roccadi 

 Papa Observatory, of which he is director, are two horizontal 

 pendulums provided with mechanical registration. It is found 

 that these instruments fail to indicate small local shocks, while 

 in recording distant earthquakes they lag behind the vertical 

 pendulums with stationary masses. But, by increasing the 



