September i6, 1915J 



NATURE 



79 



J neral statement that every symbol or sign intro- 

 Liced at a different level to the rest involves extra 

 :)acing, although this objection does riot apply to. the 

 .>e of the ordinary indices and suffixes, or to the use 

 of two lines to represent fractions in a line consisting 

 entirely of formulae. The use of a bar over a square 

 root is objectionable, but we would suggest that the 

 proposal always to use Roman or Greek i for the root 

 of — I is scarcely necessary^ as the bar can be omitted 

 in this case without ambiguity. The use of fractions 

 in the middle of a line of letterpress produces an un- 

 sightly result. While agreeing with the suggestion 

 that the small numerical fractions which are usually 

 cast as single pieces of type can often be used with 

 advantage, it should be noticed that " thirds " are not 

 easy to distinguish from "eighths," and the solidus 

 notation may often be used to overcome this difficulty. 

 The avoidance of letters with bars or dots placed above 

 them, to which much importance is attached, is a 

 change which may often present difficulty, and it is 

 doubtful how far the proposal to substitute dashes in 

 the fluxional notation will be found practical in view 

 of the frequent use of dashes for other purposes. There 

 is, however, no reason why the raised dots should 

 not be placed after instead of over the letters, while 

 inverted commas offer another new opening, hitherto 

 neglected. We should like to see "exp." included in 

 the list of suggestions. But if writers of papers are 

 sometimes to blame, printers have an awkward way of 

 beginning a formula at the end of one line of letter- 

 press and continuing it in the next line, even when 

 the author has not done so. And we think the printers 

 might at least take on themselves the responsibility 

 of removing the bars over square roots. When, how- 

 ever, a paper refers to the rigid dynamics of aeroplane 

 motions, the number of symbols required is so great 

 that violation of every rule becomes inevitable. 



In the second part of vol. iv. of the Science Reports 

 of the University of Sendai, Prof. K. Honda describes 

 a balance which allows the effect of temperature on 

 any chemical change involving loss of material by 

 evaporation or dissociation to be followed readily. 

 The beam of the balance is a silica glass tube from 

 one end of which a small platinum or magnesia vessel 

 to contain the specimen is suspended, within an elec- 

 trical heating coil enclosed in turn in a Dewar vacuum 

 vessel. The other end of the beam is pulled down by 

 a spiral spring, the pull of which can be adjusted to 

 suit the specimen used. The movement of the beam 

 is observed by means of a mirror attached to it and a 

 telescope and scale, a deflection of a millimetre corre- 

 sponding to about half a milligram. The temperature 

 of the specimen is observed by means of a thermo 

 junction, and is raised about a degree per minute. 

 Curves showing the loss of water of crystallisation 

 of two sulphates, the dissociation of a carbonate, and 

 the change of chromic oxide with temperature are 

 given. 



" Red Book " No. 198 of the British Fire Prevention 

 Committee contains the report of a fire test with three 

 window openings filled with wired glass. These were 

 subjected to the committee's standard 6o-minute test 

 at temperatures reaching up to 1600° F., followed by 

 NO. 2394, VOL. 96] 



the application of water under pressure. The usual 

 test size of glazed panels is about 2 ft. by 2 ft., but 

 one of the panels under test in this case measured 

 about 3 ft. by 2 ft. The classification of temporary 

 protection (Class B) under the committee's standards 

 has been accorded for this glazing to this larger size. 

 Partial protection (Class A) has already been obtained 

 for wired glazing by the same makers (Messrs. 

 Pilkington Brothers, Ltd., St. Helens), both for regu- 

 lation sizes and for a panel measuring about 4 ft. 

 by I ft. In the present tests, in the case of two of 

 the panels, neither fire nor water had passed through 

 the glazing at the conclusion of the test, and the 

 glazing remained in position after the appli- 

 cation of water. In the third panel (two 

 lights) fire had not passed through the glazing, but 

 on the application of water the glazing in the lower 

 light was perforated at the top portion, and some 

 water came through. In this case also th^ glazing 

 was in position after the application of water. 



Mr. James Horsfall, late chief draughtsman of 

 the Canadian Pacific Railway, suggests an improve- 

 ment in the design of locomotive coupling and con- 

 necting rods. Such rods are subject to reversal of 

 stresses due to the alternate push and pull, and also 

 in the bending due to the action of centrifugal force. 

 The effects of the latter may be reduced considerably 

 by prolonging the rods beyond the pins and attaching 

 balance weights to the extensions so formed. By 

 suitably adjusting the proportions, the bending moment 

 at the middle of the rod may be made zero, and the 

 maximum bending moment on the rod will be much 

 less than in the same rod unbalanced. Mr. Horsfall 

 says that, as the result of very numerous experiments 

 with rods of various kinds, he is convinced that the 

 best way to avoid danger from the breakage of 

 coupling and connecting rods is to counterbalance 

 them in the manner suggested. 



OUR ASTRONOMICAL COLUMN. 



Comet 19 15^ (Mellish). — The discovery of a second 

 comet during the present year by Mr. J. E. Mellish, 

 Madison, Wis., U.S.A., was announced last Thursday 

 in a telegram from Copenhagen. On September 6, at 

 22h. 91m., the comet's position was a 6h. 3o-2m., 

 S 8° 50' (approx. 8° south of 7 Geminorum). No in- 

 formation was supplied regarding the direction or rate 

 of the comet's motion, or of its magnitude. 



Saturn. — Dr. Percival Lowell, in a letter to M. 

 Flammarion (L' Astronomic, August, 1915), describes 

 observations made at Flagstaff during last spring. 

 Photographs were secured on March 12, when the ring 

 system was practically at maximum apparent breadth. 

 Cassini's division appeared continuous. With the 

 planet thus framed in its rings a new determination of 

 the aplatissement has been possible, giving the value 

 1/9-18 



Radial Velocity, Magnitude, and Spectral Type. 

 — Mr. C. D. Perrine has published (Astrophysical 

 Journal, vol. xli., pp. 396-99) the results of further 

 analysis of the data in Campbell's lists of radial 

 velocities. Dividing the stars into two groups accord- 

 ing to magnitude, drawing the line at 3.0m., he finds 

 a decided increase of the mean radial velocity for the 

 fainter group; this also holds for the stars of each 



