26 



NATURE 



[July i, 1909 



iKa, ff), x(o, 6))/(fl-o)(A-S), and <7,, ^,, . . ., aj, /3,, . . . 

 are two sets' of 2 w;? pairs of aibitrary quaniilie*. The eliminant 

 of two quanlics <J>(.t), i^^v) may be expressed as a PfaHiin 



■ 2±[i, 2][3, 4][5, 6] . . . [2«-l, 2«] 



where \i-, s\- [(^(<7'^^) i/<(«-',) - <p(d-^-^(rP-,)\l[ar-\- a,). "Xhe eliminant 

 of ifiice quanlics <))(.v, j'), ij/(.v, y), xG^, r) of the ordinary 

 standard form, 2A,.,.i 'j-', (;+jS«) is given by the Pfaltian 



whei e 



2±[l, 2][3, 4][S, 6] . . . [2«'-i,2«-i] 



(a,. - »,) («, - />r). 



— The liquidus curves of . the ternary system aluminium- 

 copper-tin : J. H. Andrew and C. A. Edwards. The 

 study of the, constitution of alloys is of great theoretical 

 interest, and of some practical value ;: in. fact, it may be 

 said that the heat treatment of a given .series of alloys 

 cannot be correctly accomplished without an accurate 

 knowledge of the structural changes which occur with 

 varying temperature and concentration. We are now in 

 possession of accurate data bearing on the constitution of 

 a large number of alloys containing only two elements, 

 but very little work has been published on mixtures of 

 three or more metals. The object of the present research 

 was to throw some light on the properties of ternary 

 alloys, and, incidentally, the effect of impurities on binarv 

 alloys. The metals from which the alloys were made had 

 the following degree of purity : — 



Per cent. 



Aluminium ... ... ... ... qq.i;? 



Copper gg.gS 



Tin gg.gS 



Freezing-point detenninatinns. — The freezing points of the 

 alloys were determined directly after mixing by means of 

 a platinum + lo per cent, iridium thermo-j unction. The 

 free ends of the wires were connected by a mirror galvano- 

 meter and balancing arrangement similar to that described 

 by Messrs. Carpenter and Keeling in their work on the 

 iron-carbon alloys. In order to locate the position of the 

 isothermal curves, more than 400 alloys and melting-point 

 determinations were made. ConcUisions. — The character of 

 the liquidus curves indicates that no well-defined ternary 

 compound is deposited from any of the liquid alloys. The 

 affinity of tin for either aluminium ' or copper' is not 

 sufficient to overcome the affinity of the last two elements 

 for each other. As a consequence of the above, curves 

 of the melting points of alloys containing a constant per- 

 centage of tin bear a striking resemblance to the liquidus 

 curve of the aluminium-copper allovs. Tin is insoluble in 

 by far the greater number of the allovs.— Studies on the 

 structure and affinities of Cretaceous plants : Dr. M. C. 

 Stopes and Dr. K. Fujii. This paper is the first account 

 to be published of the anatomy of Cretaceous plants 

 petrified in calcareous nodules. .As an introduction to the 

 flora, eighteen plants are described, all of which are new. 

 The age of these plants is Upper Cretaceous, as is deter- 

 mined from the ammonites which abound in ,the matrix 

 of the nodules, and the localitv of .all the' specimens 

 described is Hokkaido, northern- Japan. The plants 

 include one fungus, three ferns, eight gvmnosperms. and 

 six angiosperms. These numbers seem ,to represent, 

 roughly, the proportions of the flora of the, nodules as a 

 whole, of which many more specimens are in the hands 

 of the authors than are described in the present paper. 

 The most interesting of the plants are : — a new type of 

 gymnosperm, Yezonia, of which the vegetative anatomy 

 is different from that of any known genus ; a gvmno- 

 spermic fructification, also new, which there is good're.'ison 

 to believe belonged to Yezonia; an angiosperm which is 

 included in the Sabiacea? ; an angiosperm of the familv 

 Saururaceas ; and the first petrified flower, Cretovarium", 

 which has three carpels surrounded by the perianth. The 

 names of the described plants ' are : — Petrosphaeria 

 japonica. Fasciostclopteris Tansleii, Schizaeopteris meso- 

 zoica, NiponophyUum cordaitiforme, Yezonia -'iils;aris, 

 Yezostrohiis OUverii, Araiicarioxylon fanhoensis, Ccdrnxy- 

 Inn Matsinmirii, C. Yendoii, Ciinningliainiostrobiis yubari- 

 cnsis, Cryptomeriopsis antiqna, Sauniropsis niponensis, 

 NO. 2070, VOL. 81] 



Jiigloxylon Hamaoanum, Populocaiilis yczocnsis, Fagoxylon 

 hokicaidensc, Sabiocaulis Sakurii, Cretbvarinm japouicum. 

 The phylogeny and distribution of these plants is considered 

 so far as possible. 



June 17. — Sir Archibald Geikic, K.C.B., president, in 

 the chair. — The nature of the hydrogen Ifocculi on the 

 sun : Prof. G. E. Hale. Photographs of the Ha line 

 in the spectrum of the solar disc, made- on Mount Wilson 

 with high dispersion, were shown on- the screen. The 

 line appears as follows : — (1) A broad dark line, differing 

 greatly in intensity and width in- different regions of the. 

 sun. Except in eruptive or rapidly changing phenomena, 

 the differences in widlh are not. very marked. (2) Within 

 the boundaries of the dark line a narrow single or multiple 

 bright line is photographed in many parts of the sun. 

 Sometimes the appearance resembles that of the calcium 

 lines K, and Kj — i.e. the bright line , lying on its dark 

 background is divided into two components by a central, 

 dark line. In other regions the bright, line is divided into 

 a larger number of components, varying in width and 

 separation. The images of ' dark hydrogen flocculi, on 

 spectroheliograph plates taken with camera slit abou^ 

 equal in width to Ha, appear to be due, in the main, 

 to local increase in tlie intensity of the dark line. In 

 .some parts of the sun, particularly those where the line 

 is distorted, variations in the width of the line may also 

 play an important part. The increased intensity of the 

 dark line is probably the reslilt of increased absorption. 

 Slides were shown to illustrate the fact, that prominences 

 at the sun's brink are frequently recorded as dark flocculi 

 when photographed in projection against the disc. The 

 possible effects of anomalous dispersion were discussed, 

 and photographs were exhibited of the same region of 

 the sun, taken simultaneously with light from the red, 

 and violet edges of Ho. The similarity of these photo- 

 graphs apparently indicates that anomalous dispersion is, 

 not the prime factor in producing the hydrogen flocculi. 

 Certain minor differences suggest, however, that it may 

 perhaps plav a secondary part in modifying their form. — 

 The origin of certain lines in the spectrum of € Orionis 

 (Alnilani) : Sir Norman Lockyer, K.C.B., F.R.S., F. E. 

 eaxandall, and C. V. Butler. The star c. Orionis 

 (.Mnitam) is of great importance as, offering a possible 

 transition stage between the helium and bright-line stars, 

 and the only outstanding lines of unknown origin were 

 those at 4097, 4379-8, and a conspicuous double at 



I fico-S ^" ^^^ '"^^'^ °^ 4°97> the clue to the identification 

 was obtained from a spark spectrum of chromium, show- 

 ing local intensifications of certain lines at one of the 

 poles. Two of these lines were found to be the previously 

 known silicon (iv) lines, 4089, 4116, probably present as 

 impurities in the fused chromium, while one of the re- 

 maining two lines was found to coincide with the 6 Orionis 

 line at 4097. These four lines are shown under various 

 conditions in the plate, indicating the steps taken in 

 tracing their origin to nitrogen. In the spectrum of 

 nitrogen, under the special conditions which gave the 

 above lines at 4097, 4103, another line w-as found at 

 43798, which was. greatly strengthened in comparison with 

 its intensity in the ordinary spark, and this line coincides 

 with the unknown line in € Orionis. During the work 

 on the above lines, one of the photographs taken of an 

 alcohol spectrum showed abnormal intensifications on 

 either side of the oxygen line 4649-2, suggesting the 

 presence of a new double. The wave-lengths of the, com- 

 ponents of this double were determined as 4647.6,- 4650.8, 

 coinciding with the wave-lengths of the components of 

 the strong double in e Orionis. By a series of comparison 

 photographs of spectra under varied conditions, the origin 

 of the double was traced to carbon, and one of the strips 

 of the plate (carbon spark in hydrogen) shows it quite 

 isolated as it appears in the stellar spectrum. Further 

 evidence of the validitv of the identification is afforded 

 bv the peculiar nature of the components of the double. 

 — Electric induction through solid insulators : Prof. H. A. 

 Wilson. This paper contains an account of a series of 

 experiments on the variation of the capacity of ebonite 

 and other condensers, with the time of charging and with 

 the potential difference. It is shown that the capacity 



