November 25, 1920] 



NATURE 



425. 



Llnnean Society, November 4.— Dr. A. Smith Wood- 

 ward, president, in the chair. — ^J. H. Owen: Further 

 researches into the life and habits of the sparrow- 

 hawlc, Accipiter tiisus (Linn.), Pall, .\fter some pre- 

 liminary remarks on some of the less known habits 

 of the sparrow-hawk, the author showed a series of 

 nearly eighty lantern-slides depicting various incidents 

 of the incubation and nestling periods. The slides 

 were from photographs of six different nests. Of 

 special interest were series showing (i) the efforts of 

 the hen to protect the nestlings from the effects of the 

 sun, and (2) the behaviour of the hen during incuba- 

 tion as affected by climatic conditions.— H. N. Dixon : 

 The mosses of the VVollaston Expedition to Dutch 

 New Guinea. The mosses were, unfortunately, not 

 described with the higher plants, but have since been 

 worked out by the author, and have proved of great 

 interest. .Mthough consisting of only some sixty 

 gatherings, the collection contained types of at least 

 two new genera, Hymenodontopsis and Callistomium, 

 and more than a dozen new species, including two 

 new species of Dawsonia, a genus which is more 

 highly represented in New Guinea than in any other 

 part of its rather limited distribution. A further col- 

 lection bv the Rev. J. B. Clark, of the London Mis- 

 sionary Society, in the neighbourhood of Boku, British 

 New Guinea, is also included, and contains ten new 

 species, comprising a very beautiful Pterobryella, and 

 other interesting things. A small species, probably of 

 Rhizogonium, named provisionally R. orbiculare. may 

 possibly represent the ancestral form of the Rhizo- 

 goniaceae. 



Mineralogicai Society, November 9 (Anniversary 

 -Meeting). — Sir William P. Beale, Bart., president, 

 in the chair. — Dr. E. S. Simpsoa : A graphic method 

 for the comparison of minerals with four variable 

 components forming two isomorphous pairs. In the 

 spinel-chromite series the two pairs are MgO.FeO 

 and .-MjOnCrjO,, and the general formula is 

 (Mg,Fe)0.(Al,Cr),Oi. The relative molecular pre- 

 ponderances of the components of each pair stated as 

 a percentage of the maximum are given by the formulae 

 x= ioo(m-/)/(frt+/) and y = 100(0 -c)/(o+c), where 

 m, f, a, c represent the number of molecules of MgO, 

 FcO, .M,0,, and Cr,0, respectively. The values 

 of X and y, calculated from a number of 

 published analyses, and from new analyses of 

 reylonite from Camban, Western .'\ustralia, are 

 plotted on rectangular cc»-ordinates. The four 

 corners of the main square arc occupied by the 

 pure compounds MgO..M,0, (spinel), FeO.Al,Oj 

 Jheriynite), FeO.CraC (rhromite), and MgO.Cr.O, 

 (liere named pirrochromite). Sub-species and varie- 

 ties of intermediate composition are divided off in 

 symmetrical areas within the square. — L. J. Spencer: 

 Fibrolite (=:sllllmanite) as a gem-stone from Burma 

 and ("I'vlon. Water-worn, prismatic crystals from the 

 ruby mines in Upper Burma measiu-e up to ij cm. in 

 length, and are clear and transparent, with a pale 

 sapphire-blue colour and marked pleochroism. .\ 

 fine f.ircted gem cut from this material is shown in 

 the British Museum collection of minerals. Deter- 

 minations were given of the optical constants; the 

 birefringence shows a wide range, y-a being seventeen 

 times ff-a. On a somewhat similar, but etched, 

 crystal from Ceylon the axial ratios were deter- 

 mined. Other crystals from Ceylon are pale greyish- 

 green with a marke<l chatovancy.- Dr. J. W. F.vtni : 

 The origin of the alkali rock*. The alkali-igneous 

 rocks form an exceptional series varving in composi- 

 tion from arid to basic, < li.ir.irterised bv a high per- 

 centage of alkalis, especially soda, and n deficiency' 

 in alumina and the oxides of the div.ijent elements. 



They appear to occur mainly in areas where the, 

 earth's crust has, as the result of ancient folding or 

 the accumulation of granitic rocks, consolidated to 

 a considerable depth, and where the temperature 

 gradient is normally low. Such areas are rarely sub- 

 ject to new folding, but are . frequently folded, and 

 with these faults the alkali rocks appear to have a 

 genetic relation. In such areas crystallisation must 

 proceed in the sub-crustal magmas, which are 

 believed to. be basic in composition under ex- 

 ceptional pressure, with the result that minerals 

 with low specific volumes, having regard to the 

 materials of which they are composed, will pre- 

 ferentially crystallise out. Garnet, zoisite, fibrolite, 

 and kyanite are examples the materials of which 

 crystallise out under less pressure with greater 

 specific volumes. As these minerals are mainly 

 silicates of aluminium and the divalent elements, the 

 uncrystallised residue will be poor in these constituents 

 and rich in the alkalis, especially soda, which was 

 present in the original magma in greater proportion 

 than the potash. It will also contain the volatile 

 fluxes in large amount. .\s a result of the faulting 

 of the crust this residue may be pressed out, find 

 its wav upwards, and give rise bv further differentia- 

 tion to the alkali rocks. — .V. F. ttalUmond : Monticel- 

 lite from a mi.xer slag. The crystals, which are 

 essentially monticellite containing about 20 per cent, 

 of olivine in solid solution, have the following physical 

 characters : Orthorhombic, a:h: c =0-4382 : i : 0-5779 ; 

 forms 010, no, 021; refractive indices, 1-663, ''6741 

 1-680; 2V, 73J°; specific gravity, 3-20. — Dr. H. H. 

 Titomas and A. F. Hallimond : A refractometer for the 

 determination of liquid mixtures. A telescope and 

 collimator with WVbsky signal are fixed in align- 

 ment ; between them is inserted a parallel-sided 

 trough containing the liquid to be determined, in 

 which is immersed a right-angled prism of known 

 index near that of the liquid. Two images of the 

 sicnal aro formed, and the angular distance between 

 them is read on the eve.piere scale : this reading is 

 nroportional to the difference of index between the 

 ligui<l and the prism. The scale division has the snme 

 value whatever the index of the pri.sm used. 



Royal Meteorological Society, November 17. — Mr. 

 R. H. Hooker, oresident, in the chair.— C. E. P. 

 Broolci and H. W. Braby : The clash of the trades 

 in the Pacific. The mechanism of rainfall in the 

 equatorial Pacific. Considering the months January 

 to .lune, and first the area east of 180° long., the 

 trades meet at a low angle. The south-east wind is 

 the w-armer, and therefore the less dense, and rises 

 above the north-east wind on a long slant, beneath 

 which eddies are formed, giving occasional west winds. 

 From the rising south-east trade wind rain is con- 

 densed, and falls through the north-east or west winds. 

 Fast of 180° long, the trade winds meet almost at 

 right angles, and, as their densities are the same, 

 thev mingle and produce a great volume of rising air, 

 forming a low-pressure area, into which air Is drawn 

 from all sides. This low-pressure area is n mobile 

 "action centre," the position of which determines the 

 character of the season : if it lies f.ir to the west the 

 season is dry; if it takes .-in enslerlv nosition the season 

 is wet.— Dr. W. H. 8t«aveni«n : The mirage. The 

 visihilitv of the mirage was found to be dependent 

 solely on the distribution of temneralure near the 

 I'round, so determined by the altitude of the iun. 

 The appearance was, therefore, not necessarily asso- 



ciate<l with hot weather, and had been well ■ — ■• ' 



the shade temperature was lielow 50°. In 



h.id shown that the old reflection thcorv v. 



able, and Ihnt the phenomenon wna purely a refraction 



2665, VOL. 106] 



