September 27, 1917] 



NATURE 



and metres per second for wind velocity are systematic- 

 ally used, but temperature values are given in Fahren- 

 heit. Falmouth Observatory is one of the Meteoro- 

 logical Office weather stations, and the records are of 

 a high standard. 



Syn<ons's Meteorological Magazine for Septembei* 

 deals tentatively with the rainfall of August and the 

 summer of 1917. It promises later to utilise more fullv 

 the ample details supplied by numerous observers. The 

 south and south-east of England, which felt the full 

 influence of a partial drought in the middle of last June, 

 have since experienced rainstorms of considerable mag- 

 nitude. In August large portions of the country had 

 more than 6 in, of rain, the fall exceeding lo in. over 

 the elevated parts of Devon and Somerset. In the 

 normally rainy portions of England and Wales the 

 ns exceeded 20 in. during the month, and in parts 

 Ireland the measurements exceeded 10 in., in manv 

 ces constituting an extreme record for the month 

 ■er a long series of years. In Scotland the rainfall 

 IS generally less remarkable, but the mean atmo- 

 .-pheric pressure for the month was the lowest recorded 

 in August for at least fifty years. For England and 

 Wales the August rainfall was 192 per cent, of the 

 average, for Scotland 121 per cent., and for Ireland 

 192 per cent. During the wet August of 19 12 the rain- 

 fall in England and Wales was 198 per cent, of the 

 average, but in Scotland it was only 119 per cent., and 

 in Ireland 129 per cent. For the three summer months, 

 June to August, the rainfall was more than 40 per cent, 

 above the average over the greater part of the south 

 of Ireland, south-west Wales, and south-east of Eng- 

 land. Portions of Middlesex, Surrey, Sussex, and Kent 

 had an excess of more than 80 per cent, of rain, rising 

 in patches to double the average. Less than the aver- 

 age rain fell in the' north and west of Scotland, and as 

 far south as Morecambe Bay. In the northern mid- 

 lands of England the rainfall was only slightly above 

 the average. 



Scientific Paper 300 of the Bureau of Standards 

 contains a summary of the results obtained by Mr. 

 W. W. Coblentz in his researches on the emissivity 

 of the tungsten filaments of incandescent electric lamps 

 filled with nitrogen. Both straight and coiled filaments 

 have been _ investigated, the latter being the more 

 efficient owing to the diminution of the heat convection 

 and conduction losses for them. The coils had an 

 inside diameter and a pitch tAvice the diameter of the 

 filament. The radiation from an element of surface 

 within the coil was found to be 90 per cent, greater 

 and relatively redder than that from an outside 

 element. The difference between the two was found 

 to be in agreement with that calculated on the assump- 

 tion of multiple reflections within the coil. Neither 

 the internal nor external radiation E can be expressed 

 in terms of the wave-length A, and the absolute tem- 

 perature T by any formula of the Wien type, i.e. 

 EocX"''V"''2Mi^ in which c, and c. are constants. 



La Nature for September i contains an article on 

 recent improvements in wireless telegraphy by M. H. 

 Volta. Fifty lines of it have been censored by the 

 French authorities, but enough of the article remains 

 to show how well the importance of the subject is 

 appreciated in France. After describing the principles 

 on which the earlier detectors — the iron or nickel filings 

 coherer, the electrolytic, the thermoelectric, and the 

 magnetic detector operated — the author shows how 

 these have been superseded by the electronic or "valve " 

 detector, and describes several of the forms and circuits 

 at present used. By coupling a number of these 

 detectors in series a high degree of sensitiveness may 

 be secured, but Pratt, of the University of California, 

 NO. 2500, VOL. 100] 



I has introduced a combined electronic and thermo- 

 I electric detector, by means of which he has detected 

 I signals from stations 10,000 kilometres awav. The 

 General Electric Co. of Schenectady has shown how 

 electronic "valves" may be used as sources of wire- 

 less waves, and it is not too much to say that the 

 "valve" has revolutionised wireless telegraphy. The 

 problem of diminishing the disturbances which so seri- 

 ously interfere with the use of these sensitive pieces 

 of apparatus has to some extent been solved by divid- 

 ing the antenna into sections separated by inductances 

 and connected to earth by inductances which give the 

 parasitic currents a frequency differing from that of 

 the signalling current. 



The greater part of the roof of the Library of Con- 

 gress, Washington, U.S.A., is covered with tinned 

 shfiet copper, all of which is from the same manufac- 

 .turer, and was installed at the time of the completion 

 of the building in 1893-94. I'his has undergone a 

 curious corrosion process which presents some unusual 

 features, and has been investigated by P. Merica and 

 reported on in Technologic Paper No. 90 of the Bureau 

 of Standards. The roof is situated in a district un- 

 commonly free from smoke, and it is not 

 near any power station or factory produc- 

 ing smoke, so that atmospheric conditions 

 may be regarded as most unfavourable for 

 corrosion. Nevertheless, the sheet has become badly 

 pitted on the upper side, and the pits occur in general 

 along the line of surface scratches. They are appar- 

 ently unrelated to the service conditions and to the 

 direction of the rolling of the sheet. When the copper 

 became exposed, as in the present case, at the bottom 

 of the scratches on the surface, it formed a galvanic 

 couple with the alloy layer, electrolytic action set in, 

 and the copper at these points was corroded, forming 

 the pits described. This publication gives the results 

 of a study of the structure of the tin coatings on copper,, 

 and it is shown that this coating consists of at least 

 three layers, viz. a thin layer of CujSn immediately 

 next the copper, then a layer of the constituent " H," 

 containing about 60 per cent, by weight of tin, and 

 finally a layer of the eutectic of tin with copper. These 

 alloy layers are. electronegative, both to the tin and 

 the copper base. 



The seventh and concluding article of a series on 

 the evolution of the chain-track tractor appears in the 

 Engineer for September 21. In these articles, which 

 are fully illustrated, will be found a connected narra- 

 tive showing the steps by which workable machines 

 were produced. No mention is made of any work 

 done in Germany or Austria, since, so far as our con- 

 temporary is aware, these countries have played no 

 part whatever in the actual evolution of vehicles of 

 this type. Nor has France added materially to the 

 development of the chain-track mechanism. The svs- 

 tem appears to have been evolved gradually by a great 

 many workers, who have carried on their investiga- 

 tions independently, though for the most part on 

 similar lines. An exception is Mr. Diplock, who alone 

 pointed to the necessity for having the roller chain 

 flexible, and each foot of the chain track absolutely 

 free to move in any direction and within certain 

 liniits with reject to the other feet immediately 

 adjacent to it, if the best results as regards friction 

 and freedom from wear-and-tear are to be looked for. 



The latest catalogue of second-hand books (No. 169) 

 of Messrs. W. Heffer and Sons, Ltd., Cambridge, is 

 mainly devoted to school books, but there are sections 

 relating to works in botany, chemistry, geology, 

 biology, inathematics, medicine, and education, which 

 should make it useful to readers of Nature, Copies 

 of the catalogue are obtainable from the publishers 

 upon written application. 



