1446 THE PREPARATION OF PURE CARBON DIOXIDE. CLIX.-The Preparation oj' Pure Carbon Dioxide. By ROBERT CROSBIE FARMER. THE diffimlty of prepa;ring carbon dioixidel completely free1 frolm air is frequently eacoant,eireld. Nolt,wit$stlaading thah a,ll sollutlioas used for making the1 gas are prelviously boileld tlhe elliminatioln of t'he last D D traces oC air is velry diflimlh and the same1 applies tot carboln diolxide from cylinders even though moist of the gasi is bloiwn to waste1 belforehand to eliminate the air as far as possible. The difficulty is especially notiioeable in the1 testing of the stability oif gun-coltton by Will's method (Zeitsch. angew. C'hem. 1901 14 743 774), in which a rela,tivedy large quantity of carboln dioxide is used to sweep o a t the1 nitrogenous gas= and the air-correction forms a veiry high pro-portion ojf the unabsorbed gases.By the f ollloiwing simple metlhod, which haa belen useid velry satisfae torily for sobme time1 in aolnneon with tbe Will test the propolrtioln of air in carbon dioxide can be reduced ta such a degree that it is praatically immeasurable. The1 metihold depends u p n the principle that a dissodved gas can be very completely removed frolm a liquid by bubbling a selcond gas through the liquid. The carbon dioxide is prepared from solutioins of potassium hydrogen oarbolna,te and sulphurio acid each of which is freeid from air by bubbling carboin dioxide throlugh it{ belfore the ttwoi a're broughtl togethelr. The pear-shaped funnels, A A colntain potassium hydrogen caxbolnahe (300 grams to 1 litre od waf,elr) and sulphuric acid (120 0.0.to 1 litIra oC water) rmpeo-tively. Thel sohtions should be prelviously filtered throlugh glass-wooll. From the funnels thew soilutiolns pass down the1 b r a d velrti-The1 genelratolr is shown in the figure TRIPHENYLARSINE AND DIPHENYLARSENIOUS SALTS ]I 447 ca.1 tubes into1 the gelnetratling boktls. The bulk od the carbon diolxide is dra,wn olff a4t U a,nd ussd as required. A small propolrtion olf the gaa is however allowed t'o pass t,hrough the1 tube C and bubbles through t,he collumns of liquid in t'hel tlwo uprightl tubes thus remov-ing all tsa.cels of dissollve,d air. If desireld a selpara4ta solurm of carboln dioixide can be1 used for tlhis purpose). As soon as the1 air originally prelsent in t,he Woiulf e's boltltds has beisin ediminatleld the carbon dioxide! relaches a sta,te of extreme purity a.nd give@ pranti-cally no1 residue1 olf gas on absorptioln with potassium hydroxide so'lu ti on.The fit,tings of the1 Woulfel's bolt& must be absolutely frete from leaks; the b,otltlle niay if desire#d be immelrseld in water covered witah a layelr of pasa.ffin wax. The feed tlubw dip below the surfaoe of inerc$ury tlol a.vodd ba,ck-diffusioln. If the1 wholle appasa,tus is mounted on at stand it ca.n be tilted slightdy to equalise t,he back pressure in the1 two feeid tubw but. no very aacrurats balancing is neaessasy. A lit.tlle melthyl-orangel in tlhe sollutlim seirvea tlol sholw tha,t the sollutlions amre( being mixed in aboat t'he right proiportlions. When t4he tap B is turned off the intaoductioln od the reagents autIoma,ti-cally ceases. The t,aps DD which regulate the b,ubbles od gas ghauld, however be dosed when the appara.tus is not in use. The spent liquor ca,n be drawn off atl E as the fresh soJutiom. a,re addeld. The rate of eivolutiota ca.n be1 incretaseld almost in.definitslly by using a somewhat 1 a,rger ap p ar ahus. Onmi stlartad the ampparaflus need nevetr be1 discolnnected. The thanks of tlhe aut'holr are duel to the1 Director of Artillery for RESEARCH DEPARTMENT, pelrmission to publish tlhis note. ROYAL ARSENAL WOOLWICH. [Received November 3rd. 1920.