126 



NATURE 



{June 17, 1875 



usual way by means of Nicol's prisms, glass plates or other 

 appropriate polariscope. The beautiful chromatic phenomena 

 thus brought out at once indicate that amongst the causes which 

 operate to produce the hardness of glass, powerful compression 

 of the interior by the contracting exterior must be one. The 

 phenomena are, in fact, essentially those of compressed glass, 

 and the curves of colour, or black and yellow, seen when the 

 glass is examined by white or monochromatic light, indicate suc- 

 cessive curves of tension and balanced, or no-tension. In a care- 

 fully prepared glass rod of half-inch length these curves are 

 rings traversed by a well-marked black cross. In an oval the 

 rings assume the cliaracter of those seen in biaxial crystals. 

 "When plates are examined, the light being transmitted from 

 back to front, they appear to act essentially as bi-refracting plates, 

 but with crosses and bands somewhat irrCijularly distributed, 

 and capable of being referred to the angles of the plates or to 

 centres of unequal heating. 



My experiments on the mechanical properties of the glass have 

 chiefly been confined to testing its hardness and the possibility of 

 grinding it. So far as I have gone at present I make it to be 

 nearly twice as hard as ordinary glass, which it scratches with 

 ease. It can be cut with a good file well moistened wdth tur- 

 pentine, and can be ground on a stone with sand, without 

 fracturing, if great care be taken and the glass be well prepared. 

 One piece, which manifested when under the polariscope evi- 

 dences of ill-balanced tension, the neutral line lying near one 

 surface, submitted to transverse grooving, but disintegrated on 

 being ground on one surface as soon as the outer surface had 

 been ground away to near the neutral line. There appears to 

 be an easily reached limit beyond which the surfaces must not 

 be unequally removed, but as my friend Mr. Thos. Fairley, 

 F.R.S.E., has been good enough to show me, there is practically 

 no limit beyond which both surfaces may not be simultaneously 

 removed. This result, foretold by me from polariscopical analysis, 

 Mr. Fairley has kindly shown by dissolving the opposing surfaces 

 away by hydrofluoric acid. The least hard portions dissolved 

 much more readily than the thoroughly hardened, and the etched 

 surfaces show wavy lines closely following the tension lines 

 shown by the polariscope. There is further this remarkable 

 feature, that the inner portion of the glass proves to be essentially 

 common glass, which fractures in the ordinary way. Further 

 experiments are necessary for the complete elucidation of the 

 subject, and are in progress, but the preceding may be useful to 

 fellow-workers on the subject. 



Leeds, June 12 Henry Pocklington 



The House-fly— A Query 



In one of the roems in the Science Schools lately built here, 

 I have noticed, in the last week or so, great numbers of the large 

 house-fly [Musca domestica) lying dead on the floor. Last Tues- 

 day I saw one fall dead, but this is the only one. This morning 

 I counted thirty-two in a space of about three square yards. I 

 examined one under a microscope, and found that most of the 

 small hairs on its body were covered with a yellowish powder. 

 Can any of your readers give me any explanation of this ? 



Harrow, June 8 Harrovian 



OUR ASTRONOMICAL COLUMN 



Variable Stars,— Mr. J, E, Gore (Umballa, Punjab) 

 writes, under date May 5, that he believes 27 Canis Ma- 

 joris to be a variable star. It is 4 in Harding's Atlas, 

 but at present about 5| or 6, and much inferior to 28 in 

 the same constellation, which Harding rates at 5, The 

 change of brightness was first noticed in 1874, This star 

 is 4*5 in the Radcliffe Catalogues, 5 in Arg. Zones, 5^ in 

 Lacaille, and 6-5 in Heis's Catalogue ; Behrmann has 6, 

 and the lowest estimate of magnitude is 7, in Flamsteed's 

 Catalogue, with respect to which Baily remarks that there 

 is no magnitude recorded in the original observation-book, 

 and that modern observations make it 4^, Mr, Gore states 

 he has also " suspected some variation of light in the red 

 star 22 Canis Majoris (between 8 and e) ; it is usually rated 

 as of .magnitude 3 or 2,1, but for some time past it has 

 seemed rather fainter than an ordinary star of the fourth 

 magnitude," Bradley and Piazzi have this star 3-4, Flam- 

 steed, Brisbane, and Heis, 4, the Washington General 



Catalogue 5, and it is so rated once by Argelander ; in 

 Behrmann it is 4*5, 



We will take this opportunity of directing the attention 

 of our astronomical readers in the southern hemisphere 

 to Behrmann's valuable Atlas and accompanying Cata- 

 logue, which, pending the publication of Dr, Gould's 

 Cordoba " Uranometria," is the only real authority for 

 recent magnitudes of the naked-eye stars of the southern 

 heavens. It is entitled " Atlas des Stidlichen Gestirnten 

 Himmels, von Dr, Carl Behrmann" (Leipsic, 1874), and 

 contains the stars in forty-six constellations between 20° 

 of south declination and the south pole, and is arranged 

 upon the plan of Argelander's well-known work. The 

 number of stars included in the Atlas is 2,344. It was 

 formed by Behrmann in the short space of from nine 

 to ten months, beginning in the autumn of 1866, and 

 on that account, as the author remarks, there may pro- 

 bably be some omissions and errors, but it is nevertheless 

 a very meritorious and important work. It appears, from 

 Dr. Gould's report to the Minister of Public Instruction 

 of the Argentine Republic, that his " Uranometria " has 

 undergone the intended revision, and is now completed, 

 and that steps are being taken for its publication. It is 

 only one of the extensive scientific undertakings which 

 will mark the residence of this distinguished and energetic 

 astronomer at Cordoba, 



The Binary Star r^ Corona Borealis,— Mr* 

 Wilson, Temple Observatory, Rugby, has published some 

 remarks upon the tendency of recent measures of this 

 star to shorten the period ot revolution assigned by com- 

 puters hitherto, and refers to Winnecke's careful dis- 

 cussion of the measures to 1856, Winnecke's orbit, how- 

 ever, is not the latest that has been calculated, that of 

 Wijkander including measures to 1870, and the period he 

 finds, 41 '58 years, is not much different from that which 

 Mr. Wilson considers to be required by the more recent 

 measures. Still, these later observations point to a further 

 diminution of the period, the exact amount of which may 

 probably be soon determined. The following angles and 

 distances are calculated from Wijkander's orbit, and on 

 comparison of the former with the results of observation, 

 it will be found that the computed value is now about 3° 

 behind the true one, 



18720 Angle 48° -07 Distance o"'9o 

 73-0 „ 51 "98 „ o -86 



74-0 „ 56 -35 ,,0 -81 



75.0 „ 61 -32 „ o 76 



76.0 „ 67 05 „ o 70 



This orbit gives the angle too small by 5°-3 for Sir W. 

 Herschel's measure in 1782, and also too small by 4°*3 for 

 his measures in 1802, or, if these differences are expressed 

 in the form A sin. d P, - o"-o9 and - o"'04 respectively. 



Sir W. Herschel's description of his experience with 

 this star is fouad in Philosoph. Trans. 1804. On Sept. 9, 

 1 781, the position was 59° 19' n./., and on Sept, 6, 1802, 

 by "a mean of two very accurate measures" it was 

 89° 40' fi.p. (This is now found to require correction of 

 180°) Herschel further states "the distance of the two 

 stars has not been subject to any sensible alteration. 

 Sept, 9, 178 1, a very small division might be seen with 

 460. Aug, 30, 1794, they were so close that with a lo-feet 

 reflector and power of 600 a very minute division could 

 but just be perceived, April 15, 1803, with a lo-feet re- 

 flector, a very small division was also visible, with 400, 

 though better with 600. And May 15, 1803, I saw the 

 separation between the two stars with the same 7-feet 

 reflector and magnifying power of 460, with which I had 

 seen it twenty-two years before," We have from 

 Wijkander's orbit for comparison with this account : — 

 1781-69 Angle 25' -4 Position o" -98 

 1794-66 „ 80 -2 „ o -6© 



1802-68 „ 175 -5 „ o -57 



1803-37 „ 181 -6 „ o '59 



Except in 1781, it will be remarked, the distances at the 



