ndsorbuble co~npouud. Methylene-blue is suitnblc, us its ndsorptirc prolierties closely reseiiible those of brine protcins (Pig. 41%). The concentriitiori of brine protein ~'nn ~lius be eq~ri!ssed in units of un cquivnlcnt qntintity of niethylene-blue. Tlic iiccurncy of culibnitioii for protcliii or dye is ubout &6y0, giviug iin ucciirncy of &lOu/o l'or protein rekrrctl to nietliylene-blue. Tlie trullslntion of tBis cdibrution into ternis of proteiii is possible only wlieit the ninounts of other ndsorbnl)lc sribstiinces in Iiriite iiro inupprccinble. CONCI~USION J liese oliserrntioiis encoiirrige the 1)elief tlint the ~iolnrognipli woiild prove n vnliinblc tool for the inwitigtition of curing briiics nnd kindrctl fluids. Of the inetliotls of nnulysis described, tliut for nitrate is tlie inost noteworthy ns it is more nccurnto nnd rnpid tliun niiy ulrcntly nvniluble. Nitrites, nt present estininted by colorinietrie Inethods, muy be estimated far inore rupitlly with the polarogrnpli, nud with superior nccurncy, unless the concentrotion of nitrite is very low. Tlie deterininntion of proteins by the polurogruph is quicker, if less nccurnte, t linn norniul chcniicnl methods, nlid more nccurute, tlioiigli slowcr, thou estiinntions bnsed on the voluiiie of n Iirecipitnte. Chlorides are estimnted less quickly tltiiii by elcct.ronietric titrution, nltliougli more quickly thiui by t lie iisuiil titnition mcthods for wllicli proteins must be rcniowd, but the uccurncy of the polnrogrnpliic nietlioil is inferior to those based on titration. Tliese stntcnients refer to deterniinntioiis from gruplis constructed from t,liree or four observntions. Bcciirncy cnn be ini~irovetl, n t t.hc cost of increased Iubour, by the use of n lurger number of stnndnrd ndditious.
However, the highest nccurncy is frequently uiiiieces- snry in the uiinlysis of brines, and it is in tlicsc circunistunces tlint the polnrogrnph is most vnluuble. Anulyses of n brine for nll four substnnces discussed nbove cnii be cwried out in three hours, ns tlic iiistrunientol ~iiiiiiipiilti-Lions ure similar in each case. Moreover, tliis iiieuus tliut tile operutions can be cnrricd out by n relntivcly iinskilled operntive, producing per~iiunent records for subseqiicnt einminution, and this is niost convenient wltete roictitie nuulyscs must be nindc. Tile principnl obstiicle to the use of tlie instrunieiit is its high initinl cost,.
I t inay prove possible to dcterniine certuiii ot,lier w l istiiilccs iii brines by u e u ~i s of tlic polnrogriipli, SIICII its lactic ncitl, pliospllutes, untl it~e~ttngloliin. 'r!tis is ti) IJC :itteiiipted later.
The work described irbovu wns carried out us purL ol' the progrirn~ii~e of tlie Wood Investigntiou I3o;rrd, a n ~l is published I)y perttiission of tliu I~c~ii~rtnient 01' Scic!nlifia nild Iiidustriul l{esenrah. r * Plow 'l'enipernturc Heseurcli St.ntiott for l~eseurcl~ in Uiocllcniistry unil Rio~iI~ysi~s, Cumbridge IICCCIVOI web. Zn, 1011 THE ROUTINE DETECTION OF CERTAIN CANNED FOODS* SPORE=FORMINC ANABROBIC BACTERIA IN By E.*L. CROSSLEY Tlio roiitiiio esuininntioti of Inrgo llU!IlbCrd of I I U I I I ~~W I for ' ultuijrobic bnuturiit , rcqtiiros u simple, irierpenaivu uirdiuni * l{eud bufuro tlw Alicrobiolugicd Punel ul' tlie Pood Urunp on hlurcb 10, lU41..
cnpnblo of initinting rapid growth from sninl1 nnnibcra of qiorcs without tho nso of spccinl nnncrobic cultivntion, wliilst positive rcnctions ~110tild be rcndily visible nntl inrolru niiniiinm IIRC of tho microscopo for confirinntiun. Coin-Iinrntire trials with coininon atuntlnril incdin und iiietliotla IICU deacribed. None \ m a entirely sntisfuctory. IJnt n ~ncccssful nietliod \rnd evolvcd. utilising n alwcinl milk iuctlium. This metliod nlso permits n rongh cliiavilientiun of bncterinl types nnd niny be npplied to otllcr fiiotlstulF~, 1inrticolnrly h i r y iirwlactn.