354 



NATURE 



[February 13. 1902 



display of forces at ihe sun's surface. In this respect the theory 

 will besubmilted loan exhaustive test in my paper in the.-iHKa/j. 

 In one important point it involves a radical deviation from the 

 views hitherto held. So far investij^ators have almost unani- 

 mously adhered to Ihe traditional view that an increase in the 

 dynamical forces at Ihe sun's .surface indicates at Ihe .same time 

 an augmentation of his light- and heat-radiation into the 

 universe. A theory founded on this assumption would have to 

 account, not only for the extra expenditure of force into space, 

 but also for the simultaneous increa.sed development of force in 

 the sun. But in the theory here proposed the exactly opposite 

 conclusion i.s arrived at. Here the forces which we see .icting 

 on the sun are called into existence by the accumulation of such 

 parts of his radiating energy as have been frevciited from being 

 thrown off into the universe. Thus a surplus of energy 

 working on the sun means a deficit of energy communicated 

 to space. 



It will be important, then, to ascertain how far this con- 

 clusion can be verified by observed facts. Modern researches 

 seem, indeed, to corroborate this theoretical result. If the 

 theory be true, the temperature of the solar layers inside 

 the absorbing atmosphere should be higher at the maxima than 

 at the minima of solar activity, while the temperatvire of a 

 body in space, which receives its heat from the sun, should 

 vary inversely. In proof of the first conclusion I may refer 

 to Sir Norman Lockyer's results with regard to the be- 



Behrens divides the bronzes into two principal groups — those 

 rich in copper, containing from I to 25 per cent, of lin, and those 

 rich in tin, containing more than 25 per cent, of tin. With the 

 exception of the metals for mirrors (25 to 35 per cent, of tin), 

 which appear homogeneous, Mr. Behrens says that in all bronres 

 a portion rich in copper or rich in tin may be detected, forming 

 the fundamental mass, the former in alloys rich in copper, the 

 latter in those rich in tin. 



Charpy (MetatlografhisI, vol. i. p. 193) divides them into 

 those rich in copper, containing 100 to 73 per cent, of copper, 

 and those rich in tin, which are again divided into four groups — 

 o to 3 per cent, of copper, in which tin crystallises in the 

 matrix ; 3 to 55 per cent, of copper, in which a compound of tin 

 and copper crystallises out of the matrix ; 55 lu 65 per cent, of 

 copper, which have .1 structure quite homogeneous and difficult 

 to resolve ; and 65 to 73 per cent, of copper, in which hard white 

 grains crystallise in the higher eutectic. 



The curve of fusibility, as determined by Le Chatelier, is 

 composed of three branches, forming by their intersections two 

 points corresponding to alloys with ' 3 and 72 per cent, of 

 copper. 



o to 3 per cent, of copper : straight : fall. 

 3 to 72 ,, ,, uniformly curved : rise. 



72 to 100 ,, ,, almost straight : rise. 



This curve and the results of Charpy (/?«// Sue. d'Encourage- 

 ment, March, 1897) from microscopic analysis closely agree. 



. ii. R£]x>rr Mloijb Research) Comw itTcc , 



C»^fu«, MS 



V 70 6e 



fM,«t~r 



haviour of the lines widened in the spectra of sunspots, 

 fiom which he infers that the matter composing the spots must 

 be of higher temperature at the times of maxima. The second 

 conclusion, on the other hand, is corroborated by all the more 

 important researches which have recently been made regarding 

 a connection between the changes of terrestrial temperature and 

 solar activity. Of some of these I subjoin the main results 

 in I'ig. 2, which exhibits the observed changes in Ihe mean 

 annual temperature at tropical and subtropical stations and the 

 corresponding variations of solar activity. It will be seen that 

 for the whole period from 1S21 until 189S the temperature- 

 curve follows most accurately the (luclualions of the inverted 

 spot. curve, thus so far proving the validity of the second 

 conclusion, that space receives less heal at the maxima than at 

 Ihe minima of solar activity. J. Halm. 



MICROSCOPICAL EXAMINA TION OF ALIO YS 

 OF COPPER AND TIA'} 



'PllE mjcrostructure of the copper-lin alloys has been studied 

 by Behrens, Charpy, Stead and others. Recently Messrs. 

 Ileycock and Neville (f^/iit. Trans. Royal Society, 1901 ; 

 C'dasgow meeting, British Association) have published several 

 papers on the eflfecl of quenching upon the microstruclure. 



' Abstract of a paper Ijy Mr. VVillinm Campbell, Columbia University, 

 New Voik, late of the Royal School of Mines, London. Read before the 

 Institution of Mechanical Engineers on December 20, 1901. 



NO. 1685, VOL. 65] 



If, however, we study the complete cooling curve of th 

 copper-tin alloys, by Sir William Roberts-Austen (Fig. I), th 

 meaning of only a part of the curve will be found to have 

 been explained by previous workers. The branches c, d, e 

 and / remain unaccounted for. 



The result of the microscopical study of these alloys is shortly 

 as follows : — 



to I per cent. Copper. — On the addition of even 01 percent, 

 of copper to tin, a new constituent surrounding the grains of tin 

 can be seen. As the percentage of copper increases, the amount 

 of enveloping material increases also, and the tin grains decrease 

 in .size and number until about i per cent, copper: they entirely 

 disappear when the whole mass is compo.sed of the first eutectic 

 alloy. When these alloys are cast, the grains of tin are greatly 

 reduced in size. 



1 to % per cent. Copper.— When the copper is increased above 

 I per cent., thin bright needles are seen, which increase in size 

 and number and vary in their method of grouping until 8 per 

 cent, of copper is reached. Their composition varies also, 

 increasing from 33' 5 per cent. Cu to 44 per cent. Cu, as was 

 pointed out by Slead{/o«r«a/ of the Society of Chemical Industry, 

 June, 1897). Casting produces a network of fine crystallites, 

 which tend to set along definite directions forming skeleton 

 crystals. Cooling in the furnace greatly increases the size of 

 the bright crystals and diminishes their number proportionately. 



g lo ^o per cent. Copper. — With 9 per cent. Cu a new con- 

 stituent crystallises out in forms similar in section to the crystals 



