

NATURE 



[October 3, 



, inv< stigati d up to 100 I . and the 



1 inoid up to 100 1 . the results in most 



.■■■■ chei ked bj the use oi the leakage mi thod, 



the galvanometer, and the megger. Of the porcelains 



d 1 hina and Royal \\ ster wen found 



hi bi si insulatoi s. On the insulation ol some 

 ,,i the samples change ol voltage had no effect. On 

 1 ffei 1 <ii 1 ise of tempi i - .1 lai ge de- 



ol insulating power, the resistivity R following 

 closelj the law, logR n T+b, where T is tempera- 

 ture "ii the Absolute scale, and a, b are constants for 

 each sample. Erinoid is not so hygroscopic as red 

 fibre, and the resistivities of both change with tem- 

 accoi ding to th( above law. 



Iim Journal fih Gasbeleuchtung for Maj 18 last 

 -.lairs thai the I nited States produced in 1917 

 670,000 hectolitres ol " solvenaphtba," i.e. heav} 

 benzol distilling over at between 160 and 180 C, 



and from which ma} be derived 90, hectolitres 



(nr.it 1\ two million gallons) of toluene by using solve- 

 naphtha in place oi oil in water-gas plants. In an 

 inilusiri.il trial with the Lowe apparatus, where solve- 

 naphtha was substituted for gas-oil, the production of 

 carburetted gas being 14, cubic metres in twenty- 

 four hours, the expenditure of solvenaphtha was two 

 litres per cbm. of gas, the equivalent expenditures of 

 coke and steam being respectively 74 kg. and 60 kg. 

 per too cbm. gas, with a temperature of 825° C. at 

 the superheater and 15-5° C. at the condenser outlet. 

 Light oil, condensed and separated by cooling, repre- 

 sented 57 per pent, of the solvenaphtha employed, and 

 it contained 14-; per cent, of benzene and 23-7 per cent. 

 of toluene, which corresponds with 8-3 per cent, and 

 13-6 per cent, of the initial quantity of solvenaphtha. 



An account of some interesting studies on the forma- 

 tion of coke, of considerable importance from both the 

 theoretical and practical points of view, is given by G. 

 Charpy and M. Godchot in the Comptes n-mlus of the 

 Paris Vcaderm ol Sciences for Augusl 26. Starting 

 with a Brassai coal containing only 11 per cent, of 

 volatile matter and not forming a coherent coke, 

 and a Durham coal with 24 per cent, of volafile matter 

 giving alone a ven friable coke, mixtures of these two 

 were carbonised in variable proportions. The crushing 

 strength of the resulting coke varied greatly with the 

 composition of the mixture, first becoming appreciabli 

 with 20 per cent, of the Durham coal- 24 kg. per 

 sq. cm. with 25 per cent., 45 kg. per sq. cm. with 

 n : cent., So kg. per sq. cm. with ;,' per cent, and 

 then falling to zero when the proportion of- Durham coal 

 was increased by a further g per cent, to 56 per cent. 

 This rapid and unexpected variation shows the import- 

 I determining exactly by experiment the best 



proportion when carbonising mixtures of twi lis. 



Tin same Rrassac coal mixed with pitch in certain 

 proportions also furnished a hard coke, and tar could 

 replaci th pitch Results of high interest were ob- 

 tained ; i 1 cits on the effects of a preliminary 

 heating to » ature. With a Durham coal 

 containing ■■■ 1 of volatile matter, direcl car- 

 bonisation at 700 ' ■ a ■■ r a voluminous, soft, friable 

 coke, and thi he result if the coal had been 

 submitted to a pri listillation at 450 C. For 

 thirty, sixty, 01 ninety minutes before raising to 

 700 C. Hut if the preliminary distillation a1 450 ' 

 were carried out for tog, I (O, >i [50 minutes, further 

 heating at 700 C. ith strengths 

 of 41, 98, and 25 kg. pei sq, cm, vel; Pro- 

 longing the preliminat \ heal ing to 4.s<> ; 1 \ for a fui thi 1 

 m minutes 1 16-5 in all) ga 1 mlverulent 

 coke. Thus this coal, completeh coking 

 purposes in its natural state, gives a < ( normal 

 qualiu if about one-third of its vola tter is 

 1553, VOL. I02] 



1 nn.i> ed ninai 



authors poin hal this unexpected result is difl 



concile with current theories as to thi causi oi 

 coke formation. 



We have received a reprint of an articli in De 

 Natuu) b\ Mi |. W. Giltay, of Delft, di aling with 

 Mersenne and his ideas on acoustics. Marin Met 

 (1588 1648), Franciscan friar, friend of Descartes, 

 Gassendi, and rhomas Hobbes, "profond philosophe 



musicien," rdins< to Consla-ilinr 1 1 n -mi-, was, 



in the opinion of Poggendorfl ("Geschichte der 

 Physik," p. ,. ■, 1. not a physicist of the first rank, 

 but remarkable for his active correspondence, which 

 at the time replaced to some extent the scientific 

 journals ol out day. Mr. Gilta) illustrates Mersenne's 

 ideas by quotations from his two principal works, 

 " Harmonie Vniverselle" (1627) and "Traitez des 

 consonances, des dissonances, des genres, des modes 8 

 de la composition " (1635). In spite of the undeveloped 

 stati of physical knowledge, the Franciscan sometimes 



approaches nrai to our present notions, as in his 

 explanation of beats. At times he is still under the 

 influence of Aristotle, as, when discussing the pendulum 

 and the fall of bodies in vacuo, he writes, "Ton ne 

 scait si le vuide est possible, nv s'il esl quelque chose 

 de reel." His writings abound in sudden digressions 

 into theology and philosophy, but his interest in music 

 was scientific rather than artistic. 



As is generally known, in pre-war times out supplii - 



of glass apparatus for laborator\ use were largely 

 obtained from what are at prest nt • m m 



Now that we are llirown on our own resell 



manufacturers in this country have taken steps to met 1 

 the demand for such glassware. Measuring instru- 

 ments, such as burettes, pipettes, flasks, and graduated 

 cylinders, made and calibrated in this country, are 

 obtainable presumably with no more difficulty than 

 iiisr now to questions of priority as regards war 

 requirements. We have recently had an opportunity 

 of inspecting specimens of such laboratory instruments 

 manufactured by Messrs. Gallenkamp and Co., Ltd. 

 \ correspondent who has seen them remarks that the) 

 are well-made articles, and appear to he quite as 

 suitable for laboratory work as the apparatus formerly 

 obtained. The only suggestion to !"■ made is that the 

 thicker parts of the figures (enclosed between 

 lines) might with advantage be roughened, to 

 the figures more conspicuous. It may be recalled that, 

 as noted recently in these columns, arrangemi in- 

 been made by the National Physical Laboratory to test 

 and certify -Lass measuring instruments sent by manu- 

 facturers, s,, that reliance can be placed on thi accuracy 

 of am apparatus thus certified. 



In a recent issm- of the Elektrotechnische Zeitschrifl 

 it was stated thai the magnetic properties of ni 

 steel caused it to be used l>\ the (.rnn.ni Navy for 

 the construction of parts of ships near to the compass 

 in order to prevent variable effects on the compass- 

 field. Vnschiitz .<nA (Y>. now write ■> letter [Elekirot. 

 Ztschr., June 201 to stati hal this usi ol nickel-steel 

 i- by no means new, and. in fact, is a ven costly 

 method of obtaining good "compass-fields," and thai 

 the method has been almostly completely discontinued. 

 The compasses ire now almosl entire!) gyroscopic. 



The use of this type of compass has the further advan- 

 of saving large quantities of nickel, which is so 



i \|h n-iv i ■ , nil 10 obtain. 



I hi smallesi nd lightest pi at deal 1 

 testing ammeter, voltmeter, and wattmeter has been 

 constructed by the American General Electric I 

 According 1 and Western : 



trieian fjuni tg last), the ammetej and wattmetei 



