CHEMISTRY. 



practically the same as those reached when the 

 metAl was cooled in hydrogen. It was shown IM 

 the observation* tlmt tannfcen attacked the vessels 

 in win- h the determinations were made, that the 

 oxidation of t hi- tungsten was either slightly 



r that a small portion was carried mechanic- 

 ally by the wa 1 in the and 

 that the supposed trioxide of tungsten contains ni- 

 trogen, and probably hydrogen. " lu x it%w of the9e 

 facts and of the fact tlmt HKNWthM WOdetoimi- 

 nations hare been made of this oxide and nothing 

 nblished. it it evident that thr in 



rmining the ntiunic mass 

 iflfstM must be regarded as unsatisfact. 

 The atosmiewvighl of tellurium has been identic- 

 ally determinetl at 127'rt by llrauner, working with 

 lenient in combination with metals; Masumi 

 -ii:-. u-mg ^M-cimen- occurring in native sul- 

 phur of high iiurity. except for the presence of tel- 

 lurium and sole nium ; ami by Standenmaier. wnrk- 

 v an entirely different inethiMl from Brainier, 

 of tellurium in .Japan in associa- 

 um in native sulphur is noticed by 

 !uka>ige as a fact of great significance in set- 

 the place of this substance in a natural classi- 

 n <>f the elements, showing, as it does, so 

 close a habitude existing between the three ele- 



S*ott remarks tip.,n the unsatisfactory nature 

 >f the experimental evidence on which toe deter- 

 minati->ns of the atomic weight of carbon 



us determinations of the expansion pro- 

 : by the absorption of carbonic dioxide by p.it- 

 ash solutions have bean employed. When this and 

 other sources of error have oeen allowed for, the re- 

 .Ititecl values of the atomic weight of carbon are 

 6 from the combustion of carbon and 1 

 from the conversion of the monoxide into the di- 

 cing on products strictly pure and 



is free as possible from error. M. 



Wyrofiboloff and A. \' rneuil find an atomic weight 



rium. of whatever origin, very close upon 



Considering the indirect character of the 



method employed, this figure can be considered 



only as approximative to about o-2 Of 



The values Ni = 58-09 and Co = 58-93 have been 

 obtained for the atomic weights of nickel and co- 

 balt from analyses of the very carefully dried and 

 purified bromides of those metals. The separate 

 determinations are represented as having UL 

 admirably among themselves. 



l"-ing the process of the oxidation of the double 

 ammonium oxalatc. I'rof. llrauner has obtain 

 the atomic weight of thorin > = 16), a 



unrulier con-iderablv lower than th.it obtained l.v 



In their n-\ isj,,n ,f the atomic weight of magne- 

 sium. T. W. Itiehurds and H. (J. Parker, with four 

 series of determinations, get results in their final 

 nit agree surprisingly with Muri- 

 gnac's value obtained from work upon m.i_ 



and sulphate. They make the value <)= Hi, 

 Mg. = 24*962. 



Joseph Gillingham Hibbs. in his estimates of the 

 :ght of nitrogen, finds as the mean of the 

 values given by potassium and sodium nitrates 14-- 

 0003. 



The atomic weight of arsenic isd-termin. 1 by the 

 HUM observer. 11*1111: v. lij 1* peon :.; ride method 



L 74-9158. 



(Chemical Analysis. In the residual and sec- 

 ondary products resulting in the manufacture of 

 r raffinose. by crystallizing with the sugar 

 and modifying its form and increasing the ] 



on, has long caused serious and unexplainable 

 errors. Numerous methods have been proposed for 



iOQ of this substance, of several of which 

 U.I i \V. II. Haird have found a combi- 



nation that has proved -ueees>ful. From the re- 

 sult* Of its application the authors conclude that 

 iMilinose occurs in appreciable quantities m the 

 juice of n m beet. Certain peculiarities 



in the crystalli/.at ion of mix: LOrOM and 



rallintjse are pointed out in the paper of the authors, 

 \\ho find, when examining crystals under the mi- 

 croscope, that frequently solutions containing ap- 

 piirently but a verv small amount of ralVmo>e will 

 letely crystalii/e into f.n-m<> that can hardly 

 IK- distinguished from pure ra:' 1 liis modili- 



cation ,,f crystalline forms xreins to afford a ivady 

 nii-an- of detecting the presence of small amount! 

 of ra ill nose. 



A ! lent mineral ferment for the de- 



struction . matter in tnxological n 



18 described by A. \'illi-r- as produced by the aid 

 of salts of mail. L'he material toilet! 



i- i.laced in a lla^k with dilute hydrochloric acid. 

 Add a few drops of a solution of a mangane- 

 and a little nitric acid, which must be renew. 

 il become- used up. The mixture must be gently 

 heat.-d. The ga-es pn.du vboiiic acid and 



nearly pure nitrogen, and no disagreeable odors are 

 evolved. Such materials as liver, lungs, el. 



\ed in a few minute*; mn-cular liber takes 

 about an hour, and a fatty mass, seeming to contain 

 products of substitution, remains ivsjsijng the ox- 

 idi/ing aciion ,,f the mixture. 



The adoption of albumen in several induct r 

 a clarifier has brought a number of different brands 

 into the market. Some of these can be used with 



Confidence, but others are carelessly made, ale: 



adulterated. Some samples have i.een found by I'. 

 Carles to contain from IM ' <-nt. of insolu- 



ble coagulated matter, having no clarifying ] 

 gum. dextrine, and gelatin ar- d us adul- 



terants. The solution of albumen in water should 

 be transparent if free from coagulated par 1 

 When treated with tannin solution, bitartrate ..f 

 potash, and grenatine. pure albumen solution should 

 show no change. If the grenatine gives a precipi- 

 1'inin is in ex-ess. and that the albu- 

 men is Adulterated with some inert body 

 i.v.-rheated in making. If, however, the" tannin 



a precipitate in the test tub.', it proves the 



presence of gelatin in the sample. 



In samples of firedamp examined by M. II. I..- 

 Chatelier for the determination of its chemical 

 composition, the incombustil*: .,ried from 



:{! per cent, to 44'4 per cent, in volume, and con- 

 tained from per cent, to \ per cent, of carbonic 

 acid, per cent, to 0*9 per cent, of oxygen, a: 



- per cent, of i, Nitrogen is 



invariably present in <juantiti. 

 percent, to as much as :: it In al! 



the nitrogen contained argon, vary ing in propor- 

 tions from ii- 74 percent, to :t-2N per cent, of the to- 

 tal nitrogen and argon together. It was concluded 

 that argon in fire damp is not derived from the 



but is only probably present in fossil air of the 

 niferoiis 'period, cither absorbed direct 

 carried in by water. 



In a study of the verv volatile portions of j 

 leiim obtained by distilling the crude product dur- 



ne cold winter months the different hyd: 

 bons were identified by converting them into their 

 chlorine substitution " products an object which 

 was effected by bringing the vapor of the hydro- 

 carbon produ- - with chlorine. It was 

 found that the petroleum contained no normal bu- 



but iso-butane; that in isolating the octanes a 

 long series of distillations had to be carried on to 



u pure products, the octanes not beginning to 

 accumulate with any degree of purity until the 



