328 



NATURE 



IFeb. 



2, I 



therefore, in accordance with the law, repel each other but are 

 attracted by the dissimilar oscillations of other atoms. In con- 

 sequence of this repulsion they must take up the regular tetra- 

 hedral position, from which, however, under the circumstances 

 mentioned above, they may be slightly displaced. In course of 

 the discussion which ensued, Prof. Riecke followed up the sub- 

 ject with a preliminary notice of his own researches upon the 

 shape of odd-valent atoms, and showed that, in case of nitrogen, 

 phosphorus, and other pentad elements, the ether-shell is pro- 

 bably not spherical, but an ellipsoid ; three valencies being 

 situate upon the great circle at the angles of an equilateral tri- 

 angle, whilst the other two are located at the poles. If this 

 indeed be the case, the tri- and penta-valency of the nitrogen 

 group will be completely and very simply explained. 



In a paper contributed to the Royal Society of Edinburgh, 

 and to be published in a forthcoming volume of the Society's 

 Transactions, Mr. A. Crichton Mitchell gives the results of an 

 experimental inquiry, made at the Edinburgh University Physical 

 Laboratory, into the thermal conductivity of iron, copper, and 

 German silver. The method employed was that of Forbes. The 

 experiments were, for the most part, a repetition of those made 

 by Prof. Tait about ten years ago ; the main difference being that 

 the bars used were nickel-plated in order to prevent oxidation, 

 and thus render the estimation of the amount of surface loss of 

 heat a matter of greater certainty than hitherto. The results 

 arrived at confirm, in the main, those of the previous work, on 

 the same bars, already mentioned, and are of some importance 

 in deciding the question as to whether thermal conductivity in- 

 creases or decreases with increase of temperature. The principal 

 conclusion in the paper is, that in iron, thermal conductivity 

 increases with increased temperature, and that therefore iron can 

 no longer be looked upon as an exception to the rule followed 

 by the other metals, viz. that their conductivity increases with 

 temperature. 



The first rack-railway in France was opened lately at Langres 

 (in Haute-Marne), which is perched on a hill 1460 feet high, 

 and is the principal seat of the French manufacture of fine 

 cutlery. The two railway- stations which have hitherto served 

 it are several kilometres from the town, and much inconvenience 

 has been felt. The new line is made on the type of the Ri^hi 

 railway, and rises to the heart of the town. A company has 

 recently been formed to construct a funicular railway at Hong 

 Kong, for ascent of the hill overlooking the town, and much 

 frequented by the inhabitants. This line will rise to a height of 

 500 metres, and will be l6oo metres in length. The formation 

 of the ground is exclusively granite, affording a fine solid base. 

 There will be three viaducts, the largest 30 metres in length. 

 Two compound engines of forty horse-power each will be 

 installed at the upper station. 



One of the most important questions in optics is whether the 

 velocity of propagation, and therewith the wave-length of light, 

 depend on its intensity or not. A determination of wave-length 

 from prismatic decomposition is not capable of great accuracy. 

 Dr. Ebert, of Erlangen, in studying the subject, preferred a 

 more accurate method based on interference phenomena. Using 

 eight variously coloured light sources (lithium, hydrogen, 

 sodium, &c.), varied in intensity between the values I and 250, 

 he established the constancy of the wave-length and velocity of 

 propagation to nearly a millionth within those limits of intensity. 

 Considering, with these results, that the great brightness of the 

 sun does not destroy the coincidences of the Fraunhofer lines 

 with lines of our terrestrial light sources, even with the greatest 

 dispersions, Dr. Ebert thinks his affirmation (of independence of 

 intensity) is generally valid within ordinarily occurring limits of 

 brightness. 



Mr. M. W. Harrington has contributed to the American 

 Meteorological Journal for December the results of an interesting 

 inquiry as to whether the rainfall is increasing on the plains. 

 The subject is one of importance, as an annual increase of the 

 rainfall would increase the agricultural capacity of a large 

 territory. In order to come to a definite Conclusion he has used 

 two long series of observations representing the average condi- 

 tions at the epochs of 1850 and 1880. The author compares the 

 lines of equal rainfall for the two periods, and shows that, if 

 there had been an increase, any one line should have travelled 

 westwards in the interval between the two epochs. The result 

 of the inquiry shows an apparent advance along the zone in- 

 cluded between the parallels of 35° and 45°, and a regression 

 above and below these latitudes, in other words that there has 

 been apparently a consistent increase of rainfall toward the 

 plains. 



M. a. Lancaster, Meteorologist at the Royal Observatory 

 of Brussels, and Inspector of Meteorological Stations in Bel- 

 gium, has published a paper on the climate of that country in 

 1887, based on the observations at Brussels, and three other 

 stations at the west, north, and east limits of the kingdom, the 

 stations selected being typical of all other points of observation. 

 The weather for each month is discussed at considerable length, 

 and the results are compared with the normal values. The ob- 

 servations show that the mean barometer is highest in December 

 and lowest in October ; the absolute maximum temperature 

 during the year at Brussels was 9l°'o, and the minimum I5°'4. 

 The prevalent winds are from west-south-west to south-south- 

 west. The rainfall during the year amounted to 24-42 inches, 

 being considerably below the average, viz. 2878 inches. The 

 rainfall diminishes with the distance from the sea, excepting in 

 the neighbourhood of the forests ; the amount increases notably 

 in the Ardennes. 



Dr. Fines has published, in the fifteenth Bulletin Meteoro- 

 logique of the Department of the Pyrenees-Orientales, the results 

 of experiments carried on at the Perpignan Observatory : — 

 (i) To test the theory of M. Kammermann, of Geneva, for the 

 prediction of spring frosts by comparison of the readings of the 

 wet-hulb and minimum thermometers (explained in the Archives 

 dcs Sciences physiques et naturelles, vol. xiv. p. 425) ; the result 

 being that the data, so far as Perpignan is concerned, do not 

 bear out the character of accuracy attributed to them elsewhere. 

 (2) To compare the results of wind- velocity given at each instant 

 by a Bourdon's anemometer (presented to the Academy of 

 Sciences, July 30, 1882) with the records of Robinson's velocity 

 anemometer. The result of this inquiry shows a mean increase 

 in the maximum velocities of over 21 per cent, by the use of 

 Bourdon's instrument as compared with the means of the greatest 

 velocities obtained by Robinson's anemometer. 



We have received from the Imperial Observatory at Tokio 

 the monthly summaries and means of the observations of the 

 meteorological service in Japan for the year 1886, accompanied 

 by charts showing the tracks of central areas of high and low 

 barometer and by synoptic weather charts for each month. The 

 stations are supplied with good instruments (in some cases with 

 self-recording apparatus), and thee obsrvations which have now 

 been regularly taken for about eleven years under the present 

 organization compare favourably with those of European 

 systems. The climate is generally of the oceanic type — ^the 

 highest mean temperature at Tokio occurring in August. The 

 winds are governed by the monsoon seasons ; during the first 

 three months of the year the prevalent directions at Tokio are 

 north and north-west. From April to June the winds shift 

 round to the south, through east. In July and August southerly 

 winds predominate, but in September a sudden change to the 

 north occurs, and continues generally until December. Warn- 



