‘on backe Senet ode bobeRebobabotateeehuti tet 10st Beh -ded staFabaneeers 80012 Pee cher eheash andes tpeepenet Ps Waar art prge ce earete oe ieiret tebee me preeds sree Beate phen cy enncan ya repent eae paaerametae a ne nnsertyt sa an Gnartcve bias No Uet fas Setesuhetabe bene stele, ete -* r ey oe Pera eearaase Sara Rtcate ter eet Lar a lnsbrheh bulbar icaenaeteaeenle’ ae ara ecpeerertane teeet-totutateltted rene eerteesee ds IS sy sate hg etic Siesteshcaee anes ara Srednitn ashen rh leeancty ered sieatdanesessenttetesehpectetsirha eeitetee eh-Acdame thibehete tute eet tit; ss acrebeboe MekEapcustipteeeees potion kit ne Mtenesratoneier eho tie ton anti toas erete) Hr Sobeped-oe bl beasts breath abe rebyrtsy . Peat retarmosbetarerd Tei eretar Oren 0 4 Sete epeeraien ieasaeprerer eit ey Wealhte techs rive ontepen ct enact °; Maaey ee eoesre pret eee fa fishies : bebe bebod oh thee oe» Ret eneomtspametenaenar gt sean cipabacicoea rai ge bhnkeinsteoobe nennabe coeeeeeat i$ oostates olden Boost 674 pacer! anvet yieibaar 8h Petipa tegen pene sd ites resees fate Sinica ibtiantabene raibt hetiiashetee atin rete ae iat datincneie rete Baie PR atta NS anaes natasha ad aos Rosine esate Givnsocarssnerieg: beieerse vi ateutatr Rivera i Steet ~ ies a pwns. acbticcinorinas ali Wait GEE aveciaumtaiigh ri ated stat banana Bae rh telat anDe gap or aeetioe pepenpcighetint YES esto ea segechers paar presatatet posteleerenaat: Inbatmeaerow oad a ererepeny yoyo rennet + Rai Dyelt stewie ary scope or sen meg getter mtytin teenie rent eras +. ietconrcentics seater tere steerer ces ened siedsede at per he api pieer mens oer oed beeen Aint ynnly SS ele we rey ecenh With ety Mitre aot Nene we a een eseannas 014960 Sey hog ot: aud rar Tae ry idaho honing estes sereae, be-tieemecesteestits + bss Dan) fs . Ot i ertad nab bekeeabod 6 mor} eavbedrouia0, Dot venerh sp bahdsOsindesvened f aadtenern entree 3 bets ROUTE AS, stir ev sesenes Marsiicetr ae itareaceas mits hae an, fing th ak deve s}=0- i ctabe pete tateias sh Se lattes States: iehawtsensebvbonstensecheys etaisnaspeiiers ile oleielt bo eqnewceis Serres ae he bod sendehedeiebe! ark eo vet-b hon (coer pedsiceal seoes a itrtartatat atresia hatched Dracded esd cuemebemrane Shnthe pent Mey weer Diseen Rar ie arent psseegenh nsttetearrins nals esa tanta ean eget netics th o Se Hepa Soe ee eabars nat vf atta 2p) ve drernib eee a; th ress ebuga i ¢2t eae fh Fah 94 eds Pad wevetA “ in oh oboe 4 tet Poh eertedabele » AA Sasen riot Siesteate shgaarneatcnie etre eke thsint So outed i naroree bitin ped a 5 r aareetes BO Pets iaecneentemn teats crits tate rei ti dod Sa bigaet pcb tettaeed pasty sea aa estos yen oiseteeeigser G20 . eeriete Ftc Fa leet an Piette Tats peel aaberhane we Sieidociessenicicet ite hehe pees io bop beore pb hot dst Bpentensadetarb jay be ae ieasearerati ids hteleg ichaeaer being et wbute ing iehtt Paes igi Renae eon atte Sagi bed Dinsetsemeeca ges * ct ath d oh 7 eet a Se suet eral papiratesceauaimeent tea scat an DEN] patoiens sae iatatee teem rad srenmegeceri re isha Batons Rot reat PRADA atin coi shlaigsa shaken andernie horn riesehe rbd veatiorienn ies redeapenesteaupetearactea. , ehh ed oe pivoted ett bed wot o ithe Pedeb eo ave dbs pane tienen rbet bt achiel Bane Deh i 9 i sete Saatheotstee eoeae nats shrbekao sion sbeivasel isk sites hh Soares Fivieheiny vieesasttiui ates be ry A aiietts Pie Goins basket oaet of Pi sirsctererer yy ry = gonericbeie etpea see soaiee-eee essed Spepea be) sheets hed erty iene Sanbaie ech rater hiding wh-44 tt alovach Shrhese-basea-p mores bh deat Sateen Antal Oe ome beh eD ty. hat ore: nest : 104-0 DEse: wa inter Ponts! vate babel brawl clay or ble'g PRE eM at eiiehpeteseske pet Utalinahares att whtbeh ashe nee os Parts virate tiee tebe meaariaeesetd ry iaitoeegabedstsisietereh skapeegd oh SO aay Mien oy bia eh tate nepal 7 nb abe sani baness etctsd teegabeten cbt OS Sat bit tl ae aoe ; setae hioacoribieheect can vt He pt 4 = 8 so Tegebeeao ison ‘ mi) ceeaepatinat nt goats a of ey eb bbo beset shetty args hdebeiee pond a: once) i eave Giuitinass eater mt dade belt bade 5, : fedeortee bobs is a tint ate ah sont ae eeei chem) eebe Paves a sipelodrbeptyes Don) Swerrematies ways seebesd rb a abe bi barirarn Deb tee Ui Weatgten bat oP bedaded deg im tee leo eric asett shy endetot > Sitpcteesseh eg pearaeity pak whe bet at ie Ve ee treryerieu eam beh erpveet ea: oe fat: Siorwtrs atigine oye: ic ae : Siete prettch tater end eg at betel eH este i tanta afar OSltapieatled yee ets Picardtacd rgsee paosh ten ie ¥ fate! ate uuerr ee: a Lote! mth oe SP iteeibirite|acbetteentseetics vette oop" pases wand cbedegeard ps et Se tirerdens art pasa Hurhrtmsy 0 tae! Teel peewee wer pe sett vo Tasimtiegsbetepagsy aby bistae bea tee 4 Jaye bads 00.7 janie is seectchverbrae Werstorpr ner en tere “ea ita eh gt sit KT ethd Pas lel eats TTT aoa Oty rg 9 pl F bebe be os pid 4 Siberprerte erirecy9 ates ye ter epee wand onasia edocs Sa rare eit tasied peat websted spoke hah Gee pete eeke hs ~ “ rte - sie krcsceenth acta bine Weg ansted 4 se apandote coed tocaroset incase Ratios tan Shines tonne te ete enh and dea giseatassnis tito soelttnpthennrateen nefeosiane teresa pabesteriee anbistet Vleltshep oid seca faa eeihjeat pare © mest vi tener yeas bee nhabe i Oens batt ebay * inde de, Cienerirvyetinn tet) he 4 Ve aes bsbtent nieces ab ver et pod eengsncgnarioat rand nti spate sche oe Gehan hepato teeies tokdeh obeb bat aparece Nigebd a esata) Ci rb bed areca Sei car te eigd ord choy fwelyss he ctewreeie Wer reas paket frat eerie wigs fs fe Sdepeh beantea. bogehtbe boas bibeb | 3 avehbtantettaacaees homer’ tears Sara ue Robe ieee PY riats eat Fuchenby parse ot w ra cerareee bot ya A behee dey ichenenciecdsors 8 Sih chin ek gee tee RODS sire benny ees ares ‘ Seobess tentont gore Ly ogectere ity! Stbtbebeaeg ane Tears [ehaavbe Mictpertoar hepsi nev stevie) hp Thbebeostt Sivitesbearkrprasare alaca, “ seat oe xr ae eohy et . Bin penal eoranerens sa ertered et tarot! brachrlyy rye ted von A oT ested hia barocgeashomartrasieschedy ‘ myer aregyd Cat sh OTS Sadigriepes eetuir) she te Rann fats finrtasien ei ater aya Matestrasod unos neheceaise So aes Rite ae nis at fests st bitoed piciney ietieeed; Pei itreitritire peor gesrsemech Use est 8 ae Crepes CTceecmbr neat ea Ieatienhaterbaes Rice s ICAL ee reiaeatat Sabatarees setae eae sistaten at geal as CES nin ee \ihel saebe nn vost lel eacereusies bepews beatin Word saristey asp (dbeenl Ieee tpeeerenb yes ete | Pitaeesbeoreniclt ite ; oe " Ot Ae errs Vie Lat | en pitedseedne arkey: Sarlesanrin ewer uss hohe TSI Meett Cuvee re natec bangs sans aires Sh ged ah jb tape ie hs seca ae we ttt Eaehate + ate bebe Oak Pe Uh ob dng obed wet i hoor + bath Mahe Mahe hereectiarebr peceeh orice peresatet tise tet “6 iia sptens spe i eae Atel ere taegthcrdes sho Re TODes tare nares Lats sren ae tet Reasaistea Ve deed be ey bye. TESOL sseeateien ss eyat oor re sennaes, “eee Ow) ogee Bey she Racttonarete ashes rrr rey seb ager ions bs ‘bias jresesyreswsp ook ts teorinent? arb dadsh-rae8! bebids btanheoet eter ga stitbvamcbea rica: ort Saar Pet ate yey tine raceebege payne habeas ips mented vane ated epee prar} piss see) iat Pebachee je pe yyeire cha niere suite hy y Latee es ssdeg: ob lor wricnesaeg ion qcag pends ageyrrdednacae) os Hantuceses cee tales: dort terete eny ra teenie ‘t rent eae! roa hot or re erys@) avdcheandeeatt rb @ sas eben be Orson rieotinit Miata epi eh y Destsb terre pace peinesie yt ah 8-hn$ dick te seo Raper ivyreseay thd Ra : PA PA bd Reet me Ary hae tne t-8 Be tbr: + etch tee gahie-.. pba sete ny n anboa-erp 4s: pried a OE ‘ite by the Internet Archive in 2010 with funding from: jae University of Toronto A JOURNAL OF THE ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS BOARD OF EDITORS C. L. Atssrre, Chairman R. E. DoouitrLe J. P. STREET E. F. Lapp L. L. Van SLYKE VOLUME I 1915-16 Due U a ee N CAG 1916 WILLIAMS & WILKINS COMPANY BALTIMORE, MD. KS Bo ae We G 4 PG Ls f NG, on % > A c. i f Y Ne Vi } ; Rts) ere a), ‘el. er \ We Sy } ae : ay <> } Fi ANG f LVN) Ween Lm ’ | \) A! Thats a eee | - 7 fe 4 =e a ru iy i] ’ x ited ae COMPOSED AND PRINTED AT TEE WAVERLY , By THE WILLIAMS | BaLrino. PROCEEDINGS OF THE THIRTIETH ANNUAL CONVENTION. NOVEMBER 1913. CONTENTS Monpay—Mornine@ SESSION. Page Wem bers) a0 d! vASlUOLs PLCSCMUreys okt iacicisisi= - v.«:iscletetsy eeereele oie taiote elect cicie eis) stele cle oe 1 Report on phosphoric acid. By A. J. Patten and L. 8. Walker................ 8 Riepore oninitrogenky. By @ailis ELAN sas. ser sels ole oisy seetotdreecreretene ete vant sieisieses il Report on determination of potash. By H. B. McDonnell........ Bera ays fein 22 Report on availability of potash. By E. E. Vanatta......................---- 24 Determination of the availability of potash in feldspathic fertilizer by means of pot experiments. By M. F. Miller and E. E. Vanatta...................... 26 The perchlorate and gravimetric cobalti-nitrite methods for the determination OMPO CASH Bye le wb) we) ATE pee apes erveie s) winyoreccrare or eee acltate eee sie Mere 29 RepoGhon soils; BycGayon READSii. see ciecvers nit ete. oa eeeeeerereteiceercee = esi Beyer tore 33 Differences in lime requirement as indicated by the Veitch method. By A. W. laine dikes @s Miciie anys 4c ysis aveseteeiis.o 5 ara si 2 are ai siatarepe iotsreee lates oy ree renee oe 39 Method for determining the lime requirement of soils. By C. H. Jones....... 43 The effect of the presence of ammonium carbonate upon humus determinations. 1857 Vie Lely Nu Kel boty. p rate LA joallnul a eige hyopAaaianopantoncmcbc badrmb aoaoe cokes eOOeD 44 Homusidetenmination.. HB yaOn Ga Smit lsersrvse)-rertcsalsireicielsisie tees eee cinaretd eteieteia 46 Report on nitrogenous compounds in soils. By J. K. Plummer............... 49 Report on inorganic plant constituents. By W. H. McIntire and B. E. Curry.. 55 Monpay—AFTERNOON SESSION. eponmommnsecvicides. IByjs= D> Asveritit smn - srs scriceits ecies atiemies ciesieieiercie ae 59 A comparison of the iodin titration and zine chlorid methods for the analysis Gtalime-sulphurisolutionss By h..Cs ROA aa .ceatiae oece eteccresc cle ceeneiciee 76 A short method for the analysis of a lime-sulphur solution. By 8S. D. Averitt... 95 TuEsSDAY—MORNING SESSION. evonhonewabers, BV Wes OKINNEL:, «1.15 cera tamer era eels ciaaalstekehs ns 97 Report of Committee A on recommendations of referees. By B. B. Ross....... 100 Report of committee on availability of phosphoric acid in basic slag. By C. 18). \AAUTE TAG A Se RAO Oct aorias Raa p renee Rance corn nc tino VOOR aEMReHC cetida ct 102 Report of committee on food standards. By William Frear................... 108 Report of committee on editing methods of analysis. By J. K. Haywood...... 108 Report of committee on practicability of organizing for study of vegetable PLOLE Mere yo vlb a Aon cGy Kevan at, ss os, clarence ere «ies aia screener ates 109 Report on food adulteration. By Julius Hortvet...................00ceeeeees 110 Reportionycolors.) UByjVWVi.k4 Mathewson), .1.\-). 2 aclelacsa sy clele@ «/aiesieisvae eels alere ele 113 Report oumMnuitpprodueta.. sb yaks ©.) GOL. faces s a/c ened ees se eerie oe 120 Re pOny Oni winewmD yale Gr Elanbmaninyns. Sune aee camel ohne coe etie econ: 131 IRE DOLus ON DCL mE yea Grape y: aaa ois ae Lisl aeel oe ee nual ee tea Senieler io 138 Report on distilled spirits) By A. B. Adams........... Dele atta chases erate eale pati’ 143 Reporion Vinepal eb ywH nid | GOOdnOwsre ace sceee center ceceeee nonce 145 lll lV CONTENTS Page Reportiontayormnpiextracts:) WD ycA.t he rallecee cea eee eee Pesaran 146 The direct determination of volatile oil of cloves by distillation ae steam. By.JublusYHortvetiw-6 os. be oes oe lair eel eee cere ee eerie 154 President7s address.) ib yi Gals Brapsane. sesnis eee rie ene eee ereen eae eens 158 TuESDAY—AFTERNOON SESSION. Reportionsmest andshish-ss By. Were Stun eee ser ee ere eee ee eee 170 Report son fats andoils:, (By RH aWerr oe. eee)-e - oee eeeeee 181 Report on dairy, products, By JultustHonrtvet.--0-40 06... sce = eee 186 Supplemental report on dairy products. By C. H. Biesterfeld............... 194 Report/oncereallproductss, By He wie WD1Ge sma. letra neti ae 195 Report on-vegetables: ‘By i. W. Magruder: oo... nr. ri. cape sian iste ieee 199 Report on cocoa and cocoa products. By H. C. Lythgoe..................... 200 Report on tea and cofiee. By J. Me Bartlett... (025. -cm-oeieelire iil eet 203 Thersublimator anditsjuses (Biya Hehe sbi ene. epee erie teers 208 Reportionepresenvatives: ByvA Hanseeken: eo eee eee aaa 210 Report onswater i toods>) | By, Wed McGee sec ere cinerea en ee eres 218 Report on inorganic phosphorus estimation in plant and animal substances. By, Ba BS HorbesyandvAsR Ds Wiiss Owen ree ree eee eee eee 221 Report on heavy metals in foods. By H. M. Loomis......................... 244 Heavy metals in foods: tin; supplementary report. By E. L. P. Treuthardt.... 254 The determination of lead in phosphate and alum baking powders. By A. F. Seekerand Ht. Di Clayton ee... seeks eck ses else eae elem ee oer eee 264 WEDNESDAY—MORNING SESSION. Report on separation of nitrogenous bodies (meat proteins). By A. D. Emmett. 267 The estimation of glycerin in meat juices and extracts. By F. C. Cook........ 279 A study of some conditions affecting the precipitation of casein. By L. L. Van Slyke andiOS Be Winters: ac.5 sce acer a elemiee aoe eee ae ad fee cheer 281 Report of Committee C on recommendations of referees. By H. E. Barnard.. 282 Report of committee on nominations. By R. J. Davidson.................... 288 Reportionidairy products, By EM. Balleyen cess. -)-iee ee eee ee eee ee 289 Report on feeds and feeding stuffs. By W. J. Jones, Jr...................- ... 289 Report on sugar and molasses. By W.E. Cross.....................+--+0--- 314 Report on testing chemical reagents. By J. B. Rather....................... 317 Report onutannine eByiC- 1B: eBaconi pers. cecil eielestals aoe terete oh ees 329 Report of Committee B on recommendations of eee By P. F. Trowbridge 331 Report of general committee on recommendations of referees. By P. F. Trow- lsta 0 Meer ans o cic ERIM OEE oor cc aU on aee A coon oH aoST ODO STA op ONO 335 Report of committee on resolutions. By J. M. Bartlett..................-.-.- 335 Report of the auditing committee. By W. H. McIntire....................-.- 336 WEDNESDAY—AFTERNOON SESSION. Report on synthetic products. By W. O. Emery.............-..--..-+s0-0+-+> 337 Suggestions on the analysis of medicated soft drinks. By B. H. St. John..... 343 Officers and referees of the Association of Official Agricultural Chemists, 1913- CONTENTS Ve PROCEEDINGS OF THE THIRTY-FIRST ANNUAL CONVENTION, NOVEMBER 1914. Monpay—MorninG SEsSION. Wem bersranduvisttOrs pLesemumryee ace ceva ve cr= cre) taeiee lalate eiarorer=: Slkeions exes oy arel ate 353 Report on phosphoric acid. By A. J. Patten and L. S. Walker................ 360 Report on neutral ammonium citrate. By L. 8. Walker...................... 369 irrammoniumictirate.. (Bye. AceHall i i.25.c cans eee emcees viele tel scle «ss 375 Report on nitrogen. By R. N. Brackett and H. D. Haskins.................. 380 Analysis of nitrogen in leather waste. By R. P. Rose...................----- 396 Report on availability of potash. By E. EH. Vanatta........................5- 398 Report on determination of potash. By T. D. Jarrell.......................- 400 naooRy ONSOIRS Berle ews tocar pele eRe OeuO sce coon BoDansoabodcddenagn 411 A new method for the determination of lime requirements in soils. By W. H. NVECTATHIT ES yes Mest tn ie TN eae Sok a, oR INRA Aege r Daslone tne lel mbetoe elas 417 mnewinterpretation of soll analyses: By G. S: Brapsie..-cees-s ss. elas sell 418 Report on nitrogenous compounds in soils. By C. B. Lipman................. 422 Reportonvalkalijsoilss By aks Ws Eanes... so -cr cee oe eteeieal cise Joe cieae eile 424 A review and discussion of some of the methods for the determination of alkali im, Sonlisg Wl Bival Meal Ongl sharon sae ae orb pocdc CUO ROCrCmremanG doate eboctiso op msa ot 426 Monpay—AFTERNOON SESSION. HVepOLt. Ons IMsEcticides:, i Biyakir CaekCOAl Kine ery syslaicss werers roe. sveiers ie eietes 4 ele ieisveierersiers 435 Repontinoniwater:, IByaWeWelskinmer:ae. sass «conor octeine sha orc ee on senicaree 458 Report of committee on availability of phosphoric acid in basic slag. By C. B. AVA negra Se yetesstets eye evethcsiere ocape ametace sie meres ees cussofevetenesele eeVeacieeernarsrere ava ieistertaese Slew syns Steuss 461 Report of committee to codperate with other committees on food definitions. [Bay \pialllbiwesi a Orestes 26m Gide Ra Seka Heer eR io bE BASE ASTER Co mild Roto Dee ciate 462 Report of committee on the study of vegetable proteins. By T. B. Osborne.. 462 TursDAY—MoORNING SESSION. Report on food adulteration. By Julius Hortvet......................000e0e- 465 Reportioncolors:, Byn We By Mathewson's..o..0 26044 secs ceo cies cece 470 Report on saccharine products. By F. L. Shannon...................-.-- panealie, Report on! fruits and irurt products. By H.'@: Gore..-2......---.....-6-5-« 480 Reponixonrwine | By. biG Haram ann, js aaereucbeyersy=ta0s ats.cteraysfaiei clover ie eisisyecieleresele 485 Maraschinoy By J. Gakiley and As Mus Sullivsimie crs cj)-cieis © laces seer 490 Reportwounvainegar:« By) He Eis Goodnows. «. <2 aes ccc cise «ns assy essusiayn heneyets 496 Reporumonriavorneextracts«— ByeAw bs baulescnuei cs saan onieaecentscies 498 The relationship between the alcohol-soluble solids and ether-soluble solids in standard ginger extract. By C. W. Harrison and A. L. Sullivan............ 506 ReporcvOnISPlces mm yates Wie ERI GS ere rtac, lexete syayetsreyovevesete a ciais etayeteierevelefetelens Creveletarae 510 Report on baking powder. By R. E. Remington..................-.00-eee-ee 511 INEyo Oats, AHEM I] fares! NIKE, “Deedee lal, eee oooennnbennnauacacnanceuospoadodou’ 513 Presidents ACCESS wmeEs valbre wks lien Clix. verers craratercie focelaystetelaraie eraiste ereraAreis le cio EI e235 es | £3 |8535| 63] & 3 |S8s 16) o io) > = > 1) per cent |per cent |rer cent |per cent |per cent | per cent |per cent | per cent A. J. Patten, East Lansing, Mich.......] .... 118.89 | 18.65] 18.01] 18.23] 15.388 | 14.64] 15.14 W. C. Marti, East Lansing, Mich...... 0.23 18.47 | 18.74] 18.53 18.23 14.53 | 14.70] 14.65 A. K. Hart, East Lansing, Mich.. a) (0215 18.37 | 18.30} 18.30} 18.22] 14.78 | 14.47] 14.90 P. L. Hibbard, Berkeley, Calif........ tetoie 18.36 | 119.43 | 119.50 18.20 | 116.10 | 116.50 | 116.50 W. H. Dore, Berkeley, Calif........... 0.19 118.93 18.41 18.74 18.14] 15.71 | 116.65 | 116.22 L. S. Walker, Amherst, Mass. seavel |) MHS) 18.41 18.69 18.33 18.18 15.19 15.11 15.40 B. D. Wilson, Lexington, Ky. ........ 0.12 18.45 NOFA ees 18.18 15.02 15.06 15.07 E. E. Sawyer, Orono, Me.............. ae 119.28 | 118.89 | 18.41] ..... NPT Beas 14.72 J. C. Jurrjens, Madison, Wis. 0.17 18.24 DS EOU excise epes||| tected erence. |]! ayes oneal Mee B. E. Curry, Durham, N. H 0.12 119.87 | 119.81 | 119.61 | 119.93 14.94 | 116.45 14.53 Benj. Freeman, Clemson College, S.C..| 0.23 17.89 | 117.45 | 18.10 | 17.93] 14.53] 14.45] 14.69 E. G. Proulx, La Fayette, Ind......... 0.33 ASSL5a]) WSL 8u lease 18.12 | 14.77 14.83 | 14.69 W. B. Ellett, Blacksburg, Va.......... Bek 18.40] 18.18 18.15 14 :825| Sacse H. H. Hill, Blacksburg, Va............ 0.06 18.30 18.36 | 118.87 14: /64,\|| cee M. I. Watkins, Columbia, Mo soe\ Oa 119.11 STATA i esctates 15.68 | 111.26] 15.08 M. L. Lowry, Columbia, Mo..........- aes aed F Stet Freel Meee ee 14S eerie J. R. Tucker, Fayetteville, Ark........] 0.25 18.09 | 18.64 18.66 | 18.10} 14.08} 15.48 | 117.34 1 Not included in average. 1915] Phosphoric acid in Sample 1—Continued. TOTAL PHOSPHORIC ACID 2 Sees 3 ee a4 3 mois- | 3 I ae z ANALYST TURE iB a ae ES ke | EE eS os oS | o~ SS Be | Bw | Bue Be ba 54 bas 38 #5 | £2 | £35| 33 Lo) o o > per cent |per cent |per cent |per cent | per cent A. K. Burke, Geneva, N. Y........... 0.15 18.61 | 18.25 | 18.53] 18.27 H. D. Edmiston, State College, Pa.....| 0.12 18.59 | 18.63} 18.50] ..... Paul Rudnick, W. L. Latshaw, Chicago, Uris tateialelorein el oiioremie rie vie sietelsisistovetersce TOE | SSn65 18.26 | 18.46 L. B. Broughton, College Park, Md..... 18237) \|( 219520))) {8s260|)) eae A.C. Adams, College Park, Md.... ects 18215) ||) 18212'| 17.95) ) ...: S. H. Wilson, Atlanta, Ga.......... 0.21 18.05 | 18.10} 18.00} 17.90 LEO TO peo RSC EUDO DOE GOO OU aCORACROAD 0.18 18.32 18.39 | 18.31 18.17 1 Not included in average. Phosphoric acid in Sample 2. A.J, 17.41 17.25 16.86 16.77 W. C. 17.06 | 17.33 17.01 16.85 A. K. 16.96 16.66 | 16.70] 16.48 P.L. 17.37 17.03 | 117.99 16.90 W. 4H. 17.31 17.02 17.52 17.04 L. 8. 17.42 17.06 17.03 16.74 B. D. 16:84 |} 16.62) ..... 16.65 E. E. 17.59 | 17.34 W720 N Te releleiste J. G. 16.42 17/09),|) Ratecre,! |) Rees 1D don (ea saneeor sonoobee a0 7opOaonI0 0.20 117.72 | 117.60 17.35 | 118.58 Benj. 16.61 16.87 | 116.20} 16.35 E. G. 16.73 | 16.69] ..... 16.61 W. B. 16.43 16.78 RSQ aretcleits ssi 16.96 TOLORS ira ck Sc siete ste M. I. 17.41 GEO Til irereiercie 16.35 Lin Lb. La WEvirecbsegeoosteccodsocndedall jocéa. If cadod |! “Sasdo-||! coda til asaae J.R. 16.99 | 16.81] 16.7 16,55 A. K. 16.88 | 16.94] 17.19 16.80 H.D 17,11)), 16:81\)) 16:88))| 2... Paul Rudnick, W. L. Latshaw. rn 17/523) |oestener 16.72 | 17.08 PES BLOUPNCON AS pers sony ced te atsiniste tee LTC 2ONOYLT 525 |v Oko) cine... mo oo nO wi IAVETHEO See ees cole oie wiatcie's «'n'c's » melamine PATTEN AND WALKER: PHOSPHORIC ACID 11 AVAILABLE PHOSPHORIC Moly bdate method per cent ACID Volumetric method per cent 15.20 Citrate of ammonia magnesia mixture 1 Not included in average. 12 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 Phosphoric acid in Sample 3. TOTAL PHOSPHORIC ACID NYREES ARE OE ACID io] ise] 3! ao BOO NCB e | Soka ve een lia MOIS- I = = : 3 3 £4 ANALYST |) Saas a 5 Be | 8 5 | Be | BS PRS | SS |) s Bae a= a > oy se 3 s ogy Ew Ee | dws 2 wo so) 2 egg fal | 9st laser || Efe) |) cs & | ses #5 Ms) iSisra | cas ic} 3 SEE 1} 5 o} > = > 1S) per cent| per cent) per cent} per cent) per cent) per cent} per cent| per cent IMs Ab IHBAA Ot ao doshceonosacdbosoan IoC Bac 14.90 | 14.77] 14.43] 14.32] 13.43] 12.84 | 13.27 Vile (Ch WINS Saas nausoonoopopbooodosbdcs 0.47 14.61] 14.88] 14.50] 14.45] 12.95} 12.93] 12.95 IMS ESS JB EUS opanupaadpoopsbedhoadscnec 0.35 14.48 | 14.43] 14.46] 14.28] 13.31] 12.86] 13.38 P. Bee 14.88 | 14.75] 15.33] 14.75] 13.49} 13.70] 13.56 Ww. 0.33 15.02 | 14.62] 15.20] 14.50] 13.53] 13.80] 13.46 L. 0.34 14.81 | 14.82] 14.87] 14.45 | 13.17] 13.07] 13.28 B. 0.22 14.72 | 14.35] ..... 14.56 | 13.26] 13.05 | 13.09 E. S200 15°52] 14°71} 15223] ...-. TSS145 eects 13.65 AG (OR aitiys fl barrdoacpaconatssoseacoces 0.48 ree SH ee EUs | TT Socal) Beoras, Ii Noose Illesaden || oo7-o 1351 DE OP a eagonncesesacoddn so0s0 noses 0.25 | 116.37 | 15.11 } 15.07 | 116.07 | 13.07 | 114.57 | 12.52 Benj preemant. se enceeciceeses eee 0.30 14.32 | 14.06 | 14.67] 14.50] 12.69] 12.7, 12.82 Hie GegProulx estore rictiee deamehlecinieriis 0.42 HEE RE Send 14.22 | 111.94 | 111.98 | 111.89 VS sO HE So onotmadascodasons3cogs 14.46 | 14.47] 14.10] ..... Sea) Gece |) secac 1G Tee b ancosoocnnosecconnenenacs 0.22 14.73 | 14.63 | 14.63 DPS Seape |) ecco IRIE Ey stil Bagcsecen- cepasceqennpoane 0.29 15.30 | 14.57] ..... 13.51 | 111.25 | 13.38 Mela OWL ysee nee sicniteercarters Bee) | | esac (Lee eee (meceation | h wasce 12°76))|| seen J. R. Tucker 0.32 | 113.72 | 113.96 | 14.88 13.31 | 13.87 | 13.33 A. K. Burke..... 0.28 14.62] 14.60] 14.70 12.90 | 13.00 | 13.03 H. D. Edmiston............ 0.35 14.83 | 14.21 | 14.39 135 30))|esercr 12.96 Paul Rudnick, W. L. Latshaw 14.53 | ..... 14.31 13.24 | 13.27] 13.34 L. B. Broughton. 14.71 | 14.80 | 14.80 URES || coo 114.28 S. C. Adams...... oe TEES) BEA) OE SOS eee h Season || “ececc Sepion Wilson oases sete siete ersciee steleintaeis 0.28 14.46 | 14.42 | 14.02 12.74 | 12.74] 13.06 PAOLA GE verensettefeite ioe risisieiemacieiaaeielniels 0.33 14.69 | 14.58 | 14.67 | 14.45 | 13.16] 13.13] 13.19 * Not included in average. COMMENTS BY ANALYSTS. W. C. Marti: Additional work done on methods for total phosphoric acid: Two methods of solution were used: (1) Two grams of sample were dissolved in 20 ce. of hydrochloric acid and 10 cc. of nitric acid. (2) Two grams of sample were dissolved in 30 ce. of nitric acid. These solutions were diluted to 200 cc. and fil- tered and 20 cc. aliquots (equivalent to 0.2 gram of sample) were treated by the following methods: (A) Modified Sonnenschein method.—Add molybdate solution and digest as usual. Wash precipitate three times by decantation through filter with solution contain- ing 100 parts of molybdate solution, 20 parts of nitric acid (specific gravity 1.20) and 80 parts of water. Then wash with water and dissolve through filter into clean beaker. Precipitate with magnesia mixture and proceed in usual manner. (B) Modified Stéckman method.—Evaporate solution in a porcelain dish adding 5 grams of ammonium nitrate toward the end, then transfer the dish to a sand bath. Evaporate to dryness and strongly heat over naked flame so that all nitrates and organic matter are destroyed. Digest residue with fuming hydrochloric acid until 1915] PATTEN AND WALKER: PHOSPHORIC ACID 13 ferric oxid is dissolved, then dilute, filter and evaporate filtrate with nitric acid until all the hydrochloric acid is expelled. Dilute with water and proceed as under official method. (C) Regular official gravimetric method except washing yellow precipitate with ammonium nitrate solution. (D) Double precipitation method.—After first precipitation with ammonium molybdate, wash by decantation. Dissolve in ammonium hydroxid and put into digestion bath, add nitric acid until almost neutralized and then add molybdate solution. From this point proceed as usual. The results by the different methods agree very closely. All samples contained traces of iron. Total phosphoric acid in Sample 1. | SOLUTION | a (Ayia | TIBET | (Ghee | Ec Weai th et | = (ie | ee Per cent. Per cent. Per cent. | Per cent. 1 | 18.43 18.49 18.39 | 18 .35 2 ISMGG* lee 18259 18.46 P. L. Hibbard: Method 2 (A).—The Kjeldahl method gave a solution from which much silica separated on standing a few hours. Magnesia precipitates were much contaminated with both iron and silica. Method 2 (B).—Magnesia precipitates from this solution very poor in quality due to much iron and silica. Presence of extra 5 cc. of nitric acid in phosphate solu- tion improved them somewhat: Sample 1, 18.91 and 18.87; Sample 2, 17.33 and 17.44; and Sample 3, 15.21 and 15.21 per cent. Double precipitation nearly removed the iron and completely removed the silica. This is less laborious than the dehydration method (2 (B) (a) ) and far more effective in giving a pure magnesia precipitate. The operation is as follows: Decant off filtrate from the first yellow precipitate through a filter, retaining in the flaskthe yellow precipitate as much as possible. Dissolve the precipitate in ammonia, leaving solution slightly alkaline, add ammonium nitrate and water to make 100 ee. volume, warm to 65°, then add 20 ce. of molybdate solution, together with 5 cc. of nitric acid to reform the precipitate; digest as usual, and filter through same filter previously used. Proceed in ordinary way to form magnesia precipitate. Results were: Sample 1, 18.50 and 18.36; Sample 2, 17.09 and 17.25; Sample3, 14.88 and 14.83 per cent. Method 2 (B) (a)—The method for dehydration, given by the referee, is not effective and is of no use according to my experience. Dehydration in oven at 130°C. for 1 hour did not helpmuch. Results were: Sample 1, 18.86 and 18.84; Sample 2, 17.76 and 17.70; Sample 3, 15.14 and 15.02 per cent. The main source of error with these solutions seems to be the iron, which is not removed by dehydration, but is mostly avoided by double precipitation. Method 2 (B) (b).—Worked somewhat better than the aqua regia solution, prob- ably due to citrie acid present, but magnesia precipitate quite impure. Method 3 (B) (b).—Same comment as for 2 (B) (b). Method 2 (B) (c).—The citrate of ammonia magnesia mixture method seems to give good results and pure precipitates, but is objectionable on account of extra manipulation and the stirring machine required to carry it out properly. L. S. Walker: By using the (a5) method of making solutions, I found by the gravimetric method total phosphoric acid to be for Sample 1, 18.83 per cent; Sample 14 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMisTs [Vol. I, No. 1 2, 17.15 per cent; Sample 3, 14.76 per cent. I also obtained the following results on the purification of the magnesium ammonium phosphate precipitate using (a7) method of making solutions. 1. The magnesium ammonium phosphate precipitate was filtered through ashless filter paper, dried, burned off in a porcelain crucible, then dissolved in 1 to 1 hydro- chloric acid, filtered, reprecipitated and weighed in the usual manner: Sample 1, 18.11 per cent; Sample 2, 16.89 per cent; Sample 3, 14.57 per cent. 2. The magnesium ammonium phosphate precipitate was filtered through ashless filter paper, dissolved in 1 to 1 hydrochloric acid, then evaporated to dryness on a steam bath to dehydrate, afterwards dissolved in a few drops of hydrochloric acid and a small quantity of water; filtered, precipitated, and weighed in the usual man- ner: Sample 1, 17.86 per cent; Sample 2, 16.40 per cent. 3. The magnesium ammonium phosphate precipitate was treated as in 2, but after being dehydrated in beaker, was transferred to porcelain beaker with hydro- chlorie acid and water, evaporated to dryness on a steam bath and finished on a hot plate; dissolved, filtered, reprecipitated, and weighed in the usual manner: Sample 1, 18.09 per cent; Sample 2, 16.30 per cent; Sample 3, 14.47 per cent. It seems evident that the silica in the magnesium ammonium phosphate precipitate had been rendered insoluble by this method; hence a trifle lower results were obtained. By using (a4) method of solution, results by the volumetric method were as fol- lows: Sample 1, 18.75 per cent; Sample 2, 17.39 per cent; Sample 3, 15.01 per cent. B. D. Wilson: Wagner flasks were not at hand and the citrate digestion was done in graduated flasks with round bodies instead of cylindrical, having the same size neck as the Wagner flasks. J.C. Jurrjens: Much difficulty was experienced in getting rid of the silica in the case of Samples 2 and 8 with method (a7) (hydrochloric and nitric acids) and the use of ammonia and ammonium nitrate was finally abandoned in the determination of phosphoric acid by this method in these two samples. All the magnesia pre- cipitates were found to contain iron. B. E. Curry: In the case of Samples 2 and 3 by the (a7) method of making solu- tion, the precipitate of magnesium pyrophosphate contained traces of iron. The same samples also showed the presence of iron by the dehydration method. Benj. Freeman: I have followed the directions with the exception that under Total Phosphoric Acid 2 (B) (a) in dehydrating I evaporated to dryness three times instead of once. The determinations of iron and silica in the magnesia precipitates were accidentally spoiled. In the case of available phosphoric acid no iron and silica were found in the magnesia precipitates. I found that in making up the slag solutions by (as) that the addition of enough water to make the material move freely, prevented caking on addition of the sulphuric acid, and thus prevented also variation in results due to incomplete decomposition of the slag. E. G. Proulz: In the estimation of total phosphoric acid, the three methods gave concordant results. The official gravimetric method using (a,) method of making solution and the optional volumetric method were easier and shorter of manipulation. Traces of iron and silica were found in all magnesia precipitates after ignition in both total and available phosphoric acid determinations. The re- sults on available phosphoric acid seem to check very well on Samples 1 and 2 but were not concordant on Sample 3, due to the difference in the solutions as they came from the rotary apparatus and not to the three methods of analysis which in all cases gave concordant results in the same solution. When available phosphoric acid in Sample 2 solutions (a), (b), and (c) was determined in the solutions after 48 1915] PATTEN AND WALKER: PHOSPHORIC ACID 15 hours standing, the molybdate method (a) gave 15.15 per cent, the volumetric method (b), 15.12 per cent, and the citrate ammonium magnesia method (c) gave 14.90 per cent. From the average result by each method it is apparent that the molybdate method gave results between the other two methods on all three samples and the optional volumetric method gave the highest results on all three samples. The analyst is of the opinion that the traces of iron and silica found in the mag- nesia precipitates after ignition were not present in sufficient amount to affect the results materially. J. R. Tucker: Precipitates were tested after ignition for iron and silica. A trace of iron by the thiocyanate test was found in precipitates by methods 2 (A), 2 (B), and 2 (B) (a) for total phosphoric acid and by methods 3 (B) (a) and 3 (B) (b) for available phosphoric acid. No silicic acid was detected. A. K. Burke: Tested all magnesia precipitates for iron and silica, but found only the merest traces of these substances present. L. B. Broughton: The results obtained seem to indicate that it is necessary to re- move the silica (SiOz) from method (a7) before making the determination. Method (a,) seems to be preferable, as the results obtained by it compare favorably with those from (a;) with the silica removed. For the determination of the available phosphoric acid the molybdate method is to be preferred. In the citrate of am- monia magnesia method the precipitates seemed to be contaminated with an im- purity which makes it hard to filter and caused the results to be too high as compared with the molybdate method. S. H. Wilson: Instructions were followed in every instance except in making the available phosphoric acid solution. Here, an ordinary 500 cc. measuring flask was used in place of the Wagner flask advised, and instead of using the rotary apparatus the samples were shaken by hand. In every case where solutions (a,) and (a7) were used, both silicia and iron were present in the magnesia precipitates. DISCUSSION. TOTAL PHOSPHORIC ACID. The results obtained this year are encouraging but while they do not per- mit the drawing of conclusions they at least point out the direction in which the work must proceed in the future. After eliminating some results that are obviously too high or too low, the average results by the four methods show a close agreement on all three samples. The results of the indi- vidual analysts by the different methods and the results of the different analysts by the same methods, are, however, not as satisfactory, al- though the agreement is quite as good as has been obtained in the past by the methods that now 4are official for potash, nitrogen, fat, and crude fiber. It is possible that the lack of close agreement is due in part to un- familiarity with the methods on the part of some of the collaborators but they have been in use in fertilizer laboratories for a number of years, and it seems improbable that the differences can be accounted for in this way. The differences are due almost entirely to the presence of iron and silica in the magnesia precipitate. By the (as) and (a7) methods of solution, there may be both iron and silica in the magnesia precipitate 16 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 and in some cases the presence of calcium sulphate has been noted when the (as) method was used. By dehydrating the (a7) solution, the soluble silica should be removed but the iron not necessarily; consequently the variations in the result by this method are undoubtedly due to the pres- ence of iron in the final precipitate. That some of the analysts found considerable quantities of iron and silica in the magnesia precipitate while others found only traces of these sub- stances, demonstrates that more work must be done in order to determine the proper conditions that must be obtained to insure the elimination of both iron and silica at all times. In this connection it is interesting to note that in almost every instance where only traces of iron and silica were reported the results show a close agreement with the results by the volumetric method. The results obtained by the modifications reported by Mr. Hibbard of California and those by the special methods reported by Mr. Marti of Michigan are no better than the results by the regular methods. The methods reported by the associate referee, Mr. Walker, for purifying the magnesium phosphate precipitate, offer a possible means of overcoming the difficulty. Further work along these lines is necessary. The results of the individual analysts by the volumetric method show a much closer agreement than is true by the gravimetric methods. By this method the results are not influenced by the presence of iron or silica but the referee is of the opinion that it would be unwise for the association to take any action toward adopting it at this time. AVAILABLE PHOSPHORIC ACID. The results obtained for available phosphoric acid by the three methods do not show as close agreement as do the results for total phosphoric acid. The averages of the results by the three methods are fairly close but the results of the individual analysts show wide variations by all three methods. Although the molybdate method was provisionally adopted by the asso- ciation in 1911 the results by it are no better than by the other methods. The average by the volumetric method is slightly lower than by the other two but this fact has little significance since the variation in the re- sults by the different analysts is great. From the results presented it is evident that the work on methods for determining available phosphoric acid in basic slag must be continued. RECOMMENDATIONS. It is reeommended— (1) That further work be done on methods for determining total phosphorie acid in basic slag. 1915] HARE: NITROGEN 17 (2) That further work be done on methods for determining available phosphoric acid in basic slag. (3) That further attention be given to the presence of iron and silica in the magnesia precipitate and that methods designed to eliminate these substances be studied. (4) That the methods outlined for total and available phosphoric acid be tried out with a synthetic solution representing as closely as possible a solution of the average basic slag. A paper on “The use of sodium citrate for the determination of reverted phosphoric acid,’ by A. W. Bosworth, was read and published later in the Journal of Industrial and Engineering Chemistry, 1914, volume 6, number 3, page 227. A. J. Patten announced by title a paper on “A simple method for preparing neutral ammonium citrate solution,” by A. J. Patten and W. C. Marti, published in the Journal of Industrial and Engineering Chemis- try, 1913, volume 5, number 7, pages 567-68. Announcement was made by the Secretary, C. L. Alsberg, that the Secretary of Agriculture after careful consideration had decided that it was inadvisable for the Department of Agriculture to continue the pub- lication of the proceedings of this association. The reasons given were that the association should not be subsidized by the Department and that the money which the Department has spent on the proceedings could be more profitably spent on publications of its own since a pub- lication of this association would be very easily financed. The matter was referred to the executive committee for report. REPORT ON NITROGEN. By C. L. Harsg, Referee The work on nitrogen during the present year has been a continuation of that of 1912; namely, the investigation of the merits of the alkaline and neutral permanganate methods for determining the organic nitrogen activity in raw materials and mixed fertilizers, and trials of the proposed method for the estimation of nitrogen in nitrates. The instructions sent to collaborators are practically identical with those sent out in 1912. 1 Read by R. N. Brackett. 18 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMIsTS [Vol. I, No. 1 INSTRUCTIONS FOR COOPERATIVE WORK. In compliance with the recommendations for this year’s work, there have been prepared the following samples for codperative study: Sample 1. Cottonseed meal. Sample 2. Garbage tankage. Sample 3. Mixture of Samples 1 and 2 with treated leather acid phosphate and muriate of potash. Sample 4. c. p. Potassium nitrate. METHODS. The methods to be used are outlined below: (1) Determine total nitrogen in Samples 1, 2, and 3 by one of the official methods. (2) Determine available nitrogen in all samples by the alkaline and neutral permanganate methods. Alkaline permanganate method.—Transfer an amount of material equivalent to 50 mg. of water-insoluble organic nitrogen (determined by extracting 2 grams of the material on a filter paper with water at room temperature, until the filtrate amounts to about 250 cc. Determine nitrogen in the residue, making a correction for the nitrogen in the filter paper, if necessary) to a small mortar, add about 2 grams of powdered rock phosphate, mix thoroughly, transfer to a filter paper and wash with successive portions of water at room temperature until the filtrate amounts to about 250 cc. Whenmuchoilor fat is present, it is well to wash with ether before extracting with water. Dry the residue at a temperature not exceeding 80°C. and transfer from the filter to a 500 to 600 cc. Kjeldahl distillation flask (round bottom preferred, but, if flat bottom is used, incline at an angle of 30°C.). Add 20 cc. of water, 15 to 20 small glass beads to prevent bumping, and 100 cc. of alkaline permanganate solution (25 grams of pure potassium permanganate and 150 grams of sodium hydroxid, separately dissolved in water, the solutions cooled, mixed, and made to volume of 1 liter). Connect with an upright condenser to which a receiver containing stand- ard acid has been attached. Digest slowly, below distillation point, with very low flame, using coarse wire gauze and asbestos paper between flask and flame, for at least 30 minutes. Gradually raise the temperature and when danger (if any) from frothing has ceased, distill until 95 cc. of distillate is obtained, and titrate as usual. In case a tendency to froth is noted, lengthen the digestion period and no trouble will be experienced when the distillation is begun. During the digestion, gently rotate the flask occasionally, particularly if the material shows a tendency to adhere to the sides. It is recommended that as nearly as possible 90 minutes be taken for the digestion and distillation. The nitrogen thus obtained is the active water- insoluble organic nitrogen. Neutral permanganate method.—Weigh a quantity of the fertilizer, equivalent to 50 mg. of water-insoluble organic nitrogen, on a moistened 11 em. filter paper, and wash with successive portions of water at room temperature until the filtrates amount to 250 ce. Transfer insoluble residue with 25 ec. of tepid water to a 300 cc. low-form Griffin beaker, add 1 gram of sodium carbonate, mix, and add 100 ce. 0 2 per cent permanganate solution. Digest in a steam or hot-water bath for 30 minutes at the temperature of boiling water, covering the beaker with a watch glass and setting well down into the bath so that the level of the liquid in the beaker is below that of the bath. Stir twice at intervals of 10 minutes. At the end of the digestion remove from the bath, add 100 ec. of cold water and filter through a heavy 15 em. folded filter. Wash with cold water, small quantities at a time, until total filtrate 1915] HARE: NITROGEN 19 amounts to about 400 cc. Determine nitrogen in residue and filter, correcting for the nitrogen of the filter. (3) Determine nitrogen in Sample 4 as follows: To 0.5 gram of the nitrates in 600 to 700 ce. flask add 200 ce. of distilled water, 5 grams of powdered zinc, from 1 to 2 grams of ferrous sulphate, and 50 cc. of a 36° Baumé soda solution. In the neck of the flask place some glass wool and connect with the distilling apparatus. Dis- till off the ammonia and collect as usual in decinormal sulphuric acid and titrate. It is requested that results be reported on the percentage basis as follows: Total nitrogen, water-soluble organic nitrogen, active water-insoluble nitrogen, inactive water-insoluble nitrogen. RESULTS OF COLLABORATION. Sample 1. Cottonseed meal. « ACTIVE WATER 8 z INSOLUBLE ACTIVITY g g NITROGEN | & a & A a t 1 a A & & & & Zz a a qa q a q ANALYST a Bf Nc By Bo ce By 3 i=} a Bo £6 Bo Bo & ° R od od on on & z i) ao ao aod ao Z m a 2 | 5 | 2 g 3 & a m 2! « Sa Buy) |) BAN Whemearie ete (races ol inet 5 g PI EE Be a3 a8 a Ee EB < Za = a per cent | per cent | per cent | per cent | per cent |per cent | per cent G. F. Lipscomb, Clemson College, z « Chacceaho pend ee encipposee anocee .32 | 5.39 | 0.93 | 2.59 | 5.02 | 48.20] 93.30 F. N. Smalley, Savannah, Ga.......| 6.45 | 6.09 | 0.36 | 2.80 | 5.56 | 46.00] 91.30 J. B. Jackson, Auburn, Ala......... 6.40 | 5.80 | 0.60 | 2.80 | 5.21 | 48.20] 90.00 SaeadleryAuburn Alaten sce ssn 6.40 | 5.94 | 0.46 | 3.63 | 5.60 | 61.10) 94.00 @iL. Hare; Auburn, Alas... ..0..:-: 6.27 | 5.70 | 0.57 | 3.18 | 5.22 | 55.00) 91.60 JACMOUCTICNS LVLACISON Waste (ROR Oa)|) cies ia cept dienes oe i|| escuela O. F. Jensen, East Lansing, Mich...| 16.50 | 6.04 | 0.46 | 3.58 | 5.48 | 59.00] 90.80 E. F. Bailey, New Haven, Conn ...| 6.47 | 5.97 | 0.50 | 3.63 | 5.72 | 60.80] 95.80 G. L. Davis, New Haven, Conn.....| 6.49 | 5.99 | 0.50 | 3.56 | 5.79 | 59.50] 96.60 “A” Burlington, Vt..............-. 6.45 | 6.01 | 0.44 | 3.52 | 5.71 | 58-55] 95.00 Pago SUD LIN PEON) Vibis, onin So bad anoue ane ar nee ananeaben 0.0104 0.0092 JNKL Teer sed Oietato bic GPCRs a RSE SITES AEA a0 OM GOR ROT aee eee eae 0.0098 D (CaGYEInOy \iermedeco ociat Annee oanritcr 6 Or ORT eGa naan Gann AOeaoS 0.0166 0.0131 FANIGHE 305 Sooo FENG Os HORROR MOREE EERE OC COCR ae BORER EAn Orr: 0.0148 Sims (GUSVaUN ATUL) fare ateye «eee eteats aint el evra iste so.s- 5s MRNAS eveiee walqarmicingere 0.0058 0.0028 YW) ET AY a Gan SEG OO EOD TOU OC OST AOS ODEO DACoaiTGoD CuO ea rre 0.0043 AN(MOIStHMANUTE) ler. actrees leis aityals es wheal notin is Gine otere eters 0.0066 0.0065 2 ALE GRE OTE AU Ee HOR TER Caen RACE ORICe cet ont epoch 0.0066 26 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTs [Vol. I, No. 1 Potash in second washing of samples—Continued. Grams of Sample soluble potash 5. (dry soil-manureimixture) sjss0e- see sees ae ate) so doin 0.0103 0.0133 IAV ETA RC osjchiariehenns ae si F tad ated cebschal arene sie Gace chore Ghee e a cere yee 0.0118 Gudmoist;soll-manure;mixture)eeceess eae erincec eee eee 0.0160 0.0175 PAS OMA EE Nejc Ui ayo ze 5k Chg Sach ASV RE ACIS eRe ne NN CN aya CRNA ITO 0.0168 The results show that the second washing removed fully twice as much potash from the soil as was removed by the first treatment. The first treatment removed nearly all of the potash from the manures, the moist- ened manure yielding slightly less potash than the dry manure. The soil and manure mixtures gave much less potash than did the manure alone, and the dry mixture gave more than twice as much potash as the moistened mixture. The second treatment yielded considerably more potash from both the dry and moistened mixtures than from the manures alone, but not as much as that obtained from the sum of the soil and manure under the similar conditions. The sum of the two determinations as shown in the following table indicates that the manure in combination with the soil decreases the amount of water-soluble potash and the moist- ened mixture seems to retain much more of the potash than does the dry mixture. Average of combined weight of potash in both washings. Grams of Sample soluble potash Ue (AryyBOIL) Miser Nar pee eats SIO a. Ae eee OCI 0.0133 Di (TMOIBE SOL) Ie se Ne csaye repens tate, oc aee al ceese aie eens cee cel eoehehe fore aersy erase 0.0183 SiN (Gry qMan Ure) s-/-ehe. Merete ernie Sahota eee ee lat See ee 0.0932 AS (MOIS tMANUTE) erase craps arhovs ay chet eo is eacleterere eke crete use eleTonseeicParsteueicvoeie 0.0905 bu (dry,soil=-manure mixture): eesan secrete secre aeeie cee ieee 0.0703 G} (or Eon Enns GATIR) )soano peso neacencebonaeacdoecbousacece 0.0423 It is recommended by the associate referee that this study be continued another year and that the effect of decomposing green material be studied. THE DETERMINATION OF THE AVAILABILITY OF POTASH IN FELDSPATHIC FERTILIZER BY MEANS OF POT EXPERIMENTS. By M. F. Minuer and E. E. Vanarra. The experiment here reported was planned to throw light on the availa- bility of potash in feldspathic fertilizer. It was carried out partly in the small greenhouse belonging to the Entomology Department of the Uni- versity of Missouri and partly out of doors. Three-gallon stone jars were used. The soil experimented upon consisted of 4 gallons of river sand to 1915] MILLER AND VANATTA: POTASH IN FELDSPAR FERTILIZER 27 one-half gallon of silt loam, the latter being added simply to give some body to the sand but in too small a quantity to supply any great amount of potash to the growing plants. Barley was the plant used as an indicator. TRIAL I. Six pots were run in duplicate with the following treatments: Pot 1: Sodium nitrate, dried blood, and acid phosphate, plus potassium chlorid. Pot 2: Sodium nitrate, dried blood, and acid phosphate, plus potassium sulphate. Pot 8: Sodium nitrate, dried blood, and acid phosphate, plus the minimum dose of feldsphatic fertilizer. Pot 4: Sodium nitrate, dried blood, and acid phosphate, plus double the mini- mum dose of feldspathic fertilizer. Pot 5: Sodium nitrate, dried blood, and acid phosphate, plus four times the minimum dose of feldspathic fertilizer. Pot 6: Sodium nitrate, dried blood, and acid phosphate, plus six times the mini- mum application of feldspathic fertilizer. Both potassium chlorid and potassium sulphate were applied at the rate of 800 pounds per acre in the pots. The feldspar was applied in the minimum quantity, so that the soluble potash figured at 3.49 per cent would be equivalent to 50 per cent soluble potash in the application of potassium sulphate. This made the minimum application at the rate of 11,450 pounds per acre, the next 22,900, the next 45,800, and the maximum 68,700 pounds. The reason for making such heavy appli- cations is that in ordinary pot work the application of fertilizer per acre is usually doubled or tripled or quadrupled because of the small amount of feeding space for the roots. It was found that these applications of the feldspathic fertilizer were too heavy as germination was greatly interfered with in the heavier treatments. The pots were replanted a number of times, but as germination was al- ways poor it was concluded that the applications were heavy enough to be caustic and the whole experiment was repeated. TRIAL II. In the second trial the applications were made at the following rates per pot: xo. | imate | PRED BLOOD | paosmrare | ‘cHrontD | stipmate | FELDSPAR ae 1.24 2.48 6.20 2.48 Are pa ee 1.24 2.48 6.20 ata 2.48 See gels. 1.24 2.48 6.20 a 35.5 ee 1.24 2.48 6.20 71.0 Bloc 1.24 2.48 6.20 106.0 Ge. 1.24 2.48 6.20 142.0 28 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 The minimum application of feldspar here carries the same amount of soluble potash as is applied in potassium sulphate on Pot 2, figuring the amount of potash in the feldspathic fertilizer as 3.49 per cent and that in the potassium sulphate at 50 per cent. The second application of the feldspathic fertilizer would seem to be double this, the third triple and the fourth quadruple. The following table shows the rate per acre of the above applications: no. | wirmare | PRIEDBLOOD | puosrtane | “cutonrb | suLPHaTE | FELDSPAR pounds pounds pounds pounds pounds pounds Wiese 200 400 1000 400 Rake Debs 200 400 1000 vets 400 Late Saye 200 400 1000 oh cee 5725 Bs bee 200 400 1000 oon 5 ate 11450 acres 200 400 1000 att ans 17175 Goons 200 400 1000 Bric Haat 22900 The pots were watered with distilled water from time to time to main- tain the proper moisture content and after they were well started they were removed to the open air where the rainfall was allowed to supply part of the moisture needed. They were allowed to grow to full maturity with the results given below in weight of grain and straw. It was noticed that the pots giving the largest total dry weight were the last to mature, otherwise there was little difference observed in the appearance of the plants. SD DON seo See ee Stns Se “ive werent grams grams grams grams Nitrogen, phosphorus, potas- 10.2538 0.9462 eee } 17 sium chlorid 11.1837 1.0163 EP) : Nitrogen, phosphorus, potas- 13.1113 3.2887 16.4 15.7 sium sulphate 13.9187 2.1813 15.1 : WeldsparvAs.c.ccioe tater eee 6.2666 1.5334 8.8 10.8 MelAS Pare e ccs ctevielersitsiscrers ts 11.1000 1.7000 12.8 : HeldsparsBos-esec sets eerie 11.0557 0.7443 12.8 | 11.9 Heldspar Ag reas anise comers 9.8036 1.1964 11.0 : Heldspar Bijjscceiocsvessjareiicnie ee aie 8.7830 1.0170 9.8 \ 10.2 WeldspariBys. ccceene scree een ee 9.9400 1.6600 10.6 a7 Heldspar Ages tecbiece eects 10.5838 1.8162 11.4 } 126 HeldspariBae- saeeenen eee eee 11.9827 1.8173 13.8 a In the repetition of this experiment the amounts of the feldspathic fertilizer, as well as the potassium chlorid and potassium sulphate appli- cations will be still further reduced. 1915| JARRELL: METHODS FOR THE DETERMINATION OF POTASH 29 THE PERCHLORATE AND GRAVIMETRIC COBALTI-NITRITE METHODS FOR THE DETERMINATION OF POTASH. By T. D. JARRELL. PERCHLORATE METHOD. The principle of the perchlorate method is based on the fact that potassium perchlorate is insoluble in strong alcohol containing a trace of perchloric acid. The low results which the writer has found by this method in comparison with the platinie chlorid method are due, in his opinion, to the fact that potassium perchlorate is slightly soluble in the alcohol wash. Chemically pure potassium chlorid was taken as a basis to prove this fact because it is free from all foreign salts and because the potassium perchlorate would not contain any barium chlorid. A precipi- tate of potassium perchlorate weighing 0.6492 gram obtained from the c.p. salt was washed with 50 cc. of alcohol wash (95 per cent alcohol con- taining 1 ec. of perchloric acid per 200 cc. of alcohol) and finally with 10 ec. of absolute alcohol, allowing 5 minutes for the washing, and the weight was reduced 0.0086 gram or 1.4 per cent. The washing was repeated under the same conditions and the weight was again reduced at about the same ratio. Therefore, the writer believes that potassium per- chlorate is slightly soluble in the alcohol wash as outlined in the method. As the perchloric method cannot be applied in the presence of sulphates, they must be removed by the addition of sufficient barium chlorid to precipitate them. It is nearly impossible to precipitate all the sulphates as barium sulphate without getting an excess of barium chlorid and many times unless extreme care and patience are used, a large excess of barium chlorid. In that case some of the barium chlorid will not be washed out and will be weighed as potassium perchlorate. Barium chlorid is only slightly soluble in 95 per cent alcohol; 0.6528 gram of barium chlorid was weighed in a prepared Gooch crucible and washed with 50 ce. of alco- hol ‘“‘wash” as previously described and finally with 10 cc. of 95 per cent alcohol, allowing 5 minutes for the washing, reducing the original weight but 0.1482 gram, showing a residue of 0.5046 gram of barium chlorid. It is thus shown conclusively that alcohol is a very poor solvent for barium chlorid. The following table gives a comparison of the perchlorate method with the platinum method on a sample of ¢.p. potassium chlorid and several mixed fertilizers. It is seen that the perchloric method here gives results lower in every case than the platinum method. An effort was made to use but the slightest excess of barium chlorid. The only accountable reason for the low results is that some of the potassium perchlorate is dissolved by washing. 30 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMIsts [Vol. I, No. 1 Comparison of results by the perchlorate and platinum methods. POTASH B5AMPLE Perchlorate method Platinum method per cent per cent : ; 61.68 63 .04 c. p. potassium chlorid.... {fl 56 62.64 AVOCA PC Tjcneene ne cite 61.62 62.84 5.37 TEGO, cele ae en 5.66 ae 5.85 5 Average 5.63 5.78 10.21 s DaaDG er fee AEE Oe 10.00 pa 10.44 aa Average 10.22 10.84 2.30 1.92 2.92 PAS eet t oct otesdit eau 1.74 2.96 1.91 3.08 1.80 AV ETABES Weise meses kee 1.93 2.99 Even should the perchlorate method give results which agree very closely with the platinum method, the writer is unable to see any advan- tage it has over the latter method. It requires much more time to carry on a series of determinations. All of the sulphates must be removed causing an extra filtration besides the time it takes to add the barium chlorid carefully. If it were possible to find a “wash” that potassium perchlorate is not in the least soluble in and one that would completely dissolve barium chlorid and other foreign salts in the precipitate at the same time, it is believed that results could be obtained which would compare favorably with those obtained by the platinum method. GRAVIMETRIC COBALTI-NITRITE METHOD. The cobalti-nitrite method for the determination of potash in fertilizers has been investigated by the association for several years with only partial success, yet the method and results have proved encouraging and final success seems probable. The principle of the cobalti-nitrite method is based on the fact that a mixture of sodium nitrite and cobalt acetate and acetic acid made up to a specified strength, precipitates a potassium salt as di-potassium-sodium- cobalti nitrite (K,NaCo(NO.).). The cobalti-nitrite reagent is very 1915] JARRELL: METHODS FOR THE DETERMINATION OF POTASH 31 unstable; it not only decomposes on several days standing but it decom- poses during evaporation with the potassium salt. The volume of the potash solution, the volume of the reagent, the time of evaporation, and many other conditions have been found to have a tendency to produce differences in the results. The reagent when heated nearly to boiling in solution with a potash salt is partly decomposed, leay- ing a pink solution. This was especially true when the reagent was added to a large volume of potash salt, taking at least an hour for the evaporation. Most of the work noted in this paper was conducted upon c.p. potas- sium chlorid, since a method which does not work well on e.p. salts cannot be depended upon with substances of a complex nature. Some determi- nations were also made on mixed fertilizers showing a comparison of results with the cobalti-nitrite method and the platinum method. On c¢.p. potassium chlorid results ranging from about 55 per cent to 61 per cent of potash were found when the reagent was added to 50 ce. solution; the higher results (61 per cent), however, were found by adding 25 ce. of the reagent while lower results (55 per cent) were secured by adding 15 ce. of reagent. The table on page 32 shows a comparison of results obtained by the cobalti-nitrite and platinum methods on a sample of c.p. potassium chlorid and several mixed fertilizers. The results by the cobalti-nitrite method were obtained by adding 25 cc. of the reagent to 10 ce. of the potash solution and evaporating to a thick sirup. The results seem to agree in most cases fairly well with those by the platinum method. The directions sent out by the referee on potash in 1912 direct the collaborators to add about 10 cc. of the reagent to about 30 cc. of the potash solution and then to evaporate to a thick sirup. This method was tried thoroughly on several mixed fertilizers without suecess. The results were discordant, the difference sometimes being over 1 per cent on aliquots taken from the same solution. The writer believes this is due to the fact that the reagent is decomposed by continual heating, and this decom- position is hastened by dilution. It is necessary to have an excess of undecomposed reagent present at the conclusion of the evaporation in order to secure concordant results. At the conclusion a series of determinations were made on the same c.p. potassium chlorid as used at first, following as closely as possible the same conditions as followed at first, in order to learn whether after more experience with the method, results could be obtained which would agree favorably with results obtained as indicated in the table. The results found were: 61.16, 60.01, 61.71, and 61.47 per cent of potash. The average of the results obtained the first time on the c.p. potassium chlorid was 63.13 per cent of potash, agreeing very closely with the theory, while the results obtained after increased skill and familiarity with the 32 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 Comparison of results by the cobalti-nitrite and platinum methods. (Theory: Potassium chlorid = 63.17 per cent of potash.) POTASH SAMPLE Cobalti-nitrite REMARKS Cobalti-nitrite method Platinum method (Itano’s modi- method fication) per cent per cent per cent fates Aliquots 22 taken c. p. potassium chlorid: (ier Uline pooaaae fase from 2.87 | 62.64 same flask. DCU 2 BORGO ao Lon ae oe Garona iianeers: 62.84 { Aliquots oS ] 5°78 taken G5! SAREE os cicoeomUa one : Lie eC i wall from 6.20 | ; \ 5.78 manne wig flask. Averages {htkins oes oe GLiGiy | NS ee. 5.78 Aliquots 10.84 f bo 10.49 10.64 \ 10.76 | taken 25AGG Riek aon core aehacste cites 10.39 10.46 10.92 from 10.42 i Paes same 10.69 Agel AVETA COs taausiscise saietevcobvere 10.43 10.66 10.84 G re Aliquots 2:98 a0 2.92 fale DER U date tnnan Bache Ura aeoen 306 303 | 2.96 from 19° ; 3.08 same | 12.86 3.03 Anal AV ETA LCs) ao;sic c, | 1 Gill |Chamot Gill |Chamot} Gil |Chamot parts parts parts parts parts per per muon mnillcon million million million cent cent ...-| 14.52] 14.52 12.71) 13.38 mers OL|) 13:95) 9 Sas an Pale nine ee Sealers ... | 13.86} 13.54] 14.89) 97.7 | 107.5 =.) 10:35) 14.52)... ret oo eke Bere 10.71} 13.38 ...-| 10.53] 13.95) 11.40} 12.50 11.60} 12.80) 11.00) 12.50 11.40} 12.20 aarp oo| ke OO | Meer ahe au here Eas 100.00} 96.80} 97.80] 96.8 | 97.8 100.00} 96.00) 98.10} 96.0 | 98.1 50.00} 47.90) 48.00) 95.8 | 96.0 25.00] 23.10] 24.20} 92.4 |, 96.8 Sayer Et 95.3 | 97.2 A ee eLO 2 0}10) 00) ee et 90) 10.70} 11.90 11.90) 12.20 See L202 90)) 10.60), ==: Pave SGC eae aire Base ; 100 .00)105 .30|107 .20)105.3 | 107.2 ...-| 11.64] 12.58 97.4 | 101.4 10.55} 10.55 10.62) 10.55 ....| 10.58} 10.55 tee =| OL2T EGLO 41. 60 34. 90 42. 06 83. 9 101.1 41.60 40.38 41.22 97.1 99.1 300 00/174 .00|182.00] 58.0 | 60.7 1 Atter the addition of lime, soil was centrifuged instead of filtered. DETERMINATION OF NITRATES. Chamot-Pratt Modification. Preparation of the phenoldisulphonic acid reagent.—Dissolve 25 grams of pure phenol in 150 ec. of pure concentrated sulphuric acid, add 75 ce. of fuming sulphuric acid (13 per cent of SOs), stir well, and heat for 2 hours at about 100°C. 52 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMIsTs [Vol. I, No. 1 Analytical process——To 50 ce. or any convenient volume depending on the nitrate content, of the solution clarified as above, add sufficient 0.04 normal sulphuric acid to nearly neutralize the alkalinity; then if chlorids are present in appreciable amounts, a volume of nitrate-free silver sulphate solution to remove the chlorin. Heat to boiling and add a small amount of aluminum cream, filter, and wash 6 or 8 times with small amounts of hot water. Evaporate to dryness, add 2 ce. of the disulphonic reagent prepared according to the directions given by Chamot, rubbing with a glass rod to ensure complete contact. Should the residue be com- pact or vitreous due to much iron or magnesium, place on water bath for 5 minutes. Dilute with distilled water and slowly add sufficient potas- sium hydroxid solution (10 to 12 normal) until the maximum color has been produced. ‘Transfer to colorimeter cylinder, filtering if necessary, and compare with a potassium nitrate standard treated with 2 ce. of this reagent. A comparison of the results obtained with the lime method for the clarification of the suspension and those obtained with the Chamberland- Pasteur filter as the clarifying agent, would tend to show that there is a marked absorptive power possessed by the filter for nitrates, as all the determinations, except those made by Feild with the Chamot modifi- cation when 41.6 parts per million of nitric nitrogen were added, are lower than the results obtained with lime. As several of the collaborators had reported absorption in this way, the experiment was tried of run- ning a standard potassium nitrate solution through the filter and making the determinations of nitrate after passing the filter. To clean the tube 100 ce. of distilled water were forced through first. The results are given in the following table: Determination of nitrates in a standard potassium nitrate solution. PARTS PER MILLION NO. PORTION WHICH PASSED THROUGH TUBE OF NITRIC NITROGEN RECOVERED SOLUTION CONTAINING 10 PARTS PER MILLION OF NITRIC NITROGEN: Last 30 ec. portion of distilled water........ First 30 cc. of nitrate solution............... Next 25 cc. of nitrate solution..... ......... Next 50 ec. of nitrate solution.......... Poses Next 75 cc. of nitrate solution............... Next 200 cc. of nitrate solution.............. Ore WN ANAWMWO CUDnDaroO SOLUTION CONTAINING 100 PARTS PER MILLION OF NITRIC NITROGEN: Last 30 ec. of distilled water................ First 25 cc. of nitrate solution............... Next 25 cc. of nitrate solution..... ......... Next 50 cc. of nitrate solution............... Next 75 cc. of nitrate solution............... Next 180 ec. of nitrate solution.............. io }h. for) wow BARES SO OCwrta- oO Dore Whe 1915] PLUMMER: NITROGENOUS COMPOUNDS IN SOILS 53 The use of aluminum cream is questionable as shown in the following results on a sample of soil sent out: Determination of nitrates by Chamot-Pratt method. NITRIC NITROGEN ANALYST With aluminum cream Without aluminum cream i parts per million parts per million CEB LAMAN castes cles: 11.12 12.00 10.56 12.00 10.08 12.20 10.88 12.00 PAW OT AOR RAM es ag claies a ellateyeee 10.66 12.05 dig LGA JEW nS eens so ooapaee 11.80 12.20 12.00 11.85 11.60 12.40 11.00 12.00 PAVIOT ARES sisrerstapaisercl tone eis 11.60 12.11 COMMENTS BY COLLABORATORS. C. B. Iipman: For the clarification of soil extracts, I am inclined to the use of the lime method without the carbon dioxid step. The excess of lime causes little or no trouble and its removal with carbon dioxid causes the loss of time and may give rise to a loss of nitrate. In some preliminary work with the Chamberland-Pasteur filter losses of nitrate were observed when the whole 200 cc. of suspension were forced through the tube. When 2 cc. of the freshly-prepared disulphonic acid under Gill’s directions are used, the results are about the same as with the Chamot reagent. The latter, however, without aluminum cream, gives higher results. It entails more trouble in its preparation. A. L. Feild: While the Chamot reagent is undoubtedly more scientifically pre- pared and gives more uniform colors to match with the standard, the Gill method offers some advantages. With the latter method, less reagent is used, consequently a smaller volume of the yellow solution can be obtained, and in soils having a low nitrate content more satisfactory readings can be obtained. G. W. Walker: It appears to me that the Gill method of procedure in the analyt- ical process is better, asit gives about as good results and is shorter. The Chamot method for the preparation seems to give a reagent of more uniform color than the Gill. L. L. LaShell: The Gill method is more easily carried out at the point of develop- ing color, as potassium hydroxid gives crystals which separate out and larger volumes of yellow solution must be taken for comparison. This will make it diffi- cult to read soils which have a low nitrate content. Volume of solution is too small to permit filtering through Chamberland-Pasteur filter. C. J. Schollenberger: The Chamot modification appears to give higher and more consistent results than the Gill, but it will be difficult to read with soils low in ni- trate, due to volume of solution necessary to dissolve the crystals which separate on the addition of potassium hydroxid. Five-tenths of a gram of lime is not enough to coagulate the clay content of this sample; it requires from 1 to 2 grams. Boil- ing off the excess of carbon dioxid causes solution of some of the humus, giving 54 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 foreign color. Removal of excess of lime is not necessary. After the addition of lime the centrifuge offers a good means of removing the clay. The Chamberland- Pasteur filter absorbs too large a proportion of nitrate to allow its use for this purpose. H.C. McLean: I am inclined to use the Gill method of procedure, as it is shorter and gives about as good results. The color produced by the Chamot reagent has a more uniform color to deal with than that of Gill’s reagent. CONCLUSIONS. It would seem from a comparison of the results that lime is the best agent known for the clarification of soil extracts in which the nitrate de- terminations are to be made. The absorptive power possessed by the Chamberland-Pasteur filter tube would tend to debar this piece of appa- ratus for this work. The amount of lime necessary for the removal of clay varies in different soils, but 2 grams are about all that are needed for 100 grams of most loam soils. It can be added to the soil itself, so as to have it present in the shaking process, or better after the shaking, by drawing off the sus- pension and adding the lime to this, shaking at intervals for 30 minutes. The Chamot modification, except in two instances (Walker and Feild), gave higher results than the Gill method on the soil alone, and in every instance in which potassium nitrate was added. The agreement between the analysts on the soil alone was remarkably close when lime was used. When the Chamberland-Pasteur filter was used for the clarification there was not such agreement. While Feild’s results on the soil alone were in fair agreement, they were lower than the average for the lime method, and his results after the addition of potassium nitrate are decidedly varying. The results obtained by McLean are much lower than the lime method. He says, however, that he did not shake continuously, but at intervals, which may be the cause of the variance in his results. The use of aluminum cream as recommended by Chamot is questionable, but the fact that this method gives higher results in every instance in which potassium nitrate was added, and in every instance except two on the soil alone is significant. That this reagent is more scientifically pre- pared and that the color produced from it is more satisfactory to match with the standard can hardly be disputed. RECOMMENDATIONS. It is reeommended— (1) That the reduction method for the determination of nitrates in soils be studied, as the colorimetric method is worthless in the case of ‘alkali soils.” (2) That the colorimetric methods for the estimation of nitrites and ammonia be studied. 1915] McINTIRE AND CURRY: INORGANIC PLANT CONSTITUENTS 55 REPORT ON INORGANIC PLANT CONSTITUENTS. By W. H. McIntire, Referee, and B. E. Curry, Associate Referee. The association at its 1912 meeting directed a further study of the Schreiber method for total sulphur as given in Bureau of Chemistry Cir- cular 56, and additional study of its official method for ferric oxid and aluminum oxid as extended to the determination of calcium and magnesium. The work on sulphur was done under the direction of the associate referee, to whom credit is due for the work here reported upon samples of bran and linseed meal. COMPARISON OF THE SCHREIBER AND SODIUM PEROXID METHODS FOR TOTAL SULPHUR. Determination of total sulphur in bran and linseed meal. TOTAL SULPHUR METHOD AND ANALYST Bran Linseed meal Schreiber method: per cent per cent Firman Thompson, Newark, Del............. 0.754 1.026 0.541 1.103 0.514 1.011 0.739 1.131 0.611 1.046 OS7O9FRGe: itty oS: PAS EN ADORNS teres Tate Anette Asraie eormiciits Sec eM 0.645 1.063 Maximum vaniatlonia.n ess. eet nein cence 0.240 0.120 G) Hy Boltzs Wooster, Ohio. 2.2. sace aces. so 0.711 1.029 0.691 1.033 0.707 1.041 0.684 1.029 0.679 1.061 OR692- is iy Pelee orice JAN OLE; Aas DOP ISIN Ee DIRE GOOCH OEIC Oe eae: 0.694 1.039 VEAXIMLUMISVATLAtLON «cys yes ccceile olsetel seis 0.032 0.032 H. Rosenthal, Columbia, Mo................ 0.746 1.151 0.634 1.116 OL634: hip itn © ashen: ONG74 8 Le ees PACU GLA Gey Rtas Sie sick ccnartipetehais, te ls eieue sisi aval 0.672 1.134 0.112 0.035 ~ A. J. Patten, East Lansing, Mich...... ..... 0.713 1.114 0.706 1.084 BAN CVAD Ont ree carta can ncas tts anal ua ecto istelasauatal es 0.710 1.099 Miaximuniivantations-2 asc) scie tee cecsie~ 0.007 0.030 Granda eradoe rennet crt type aon cilevres 0.680 1.084 Sodium peroxid (official) method: | ..... 1.063 etic 1.108 PAV CLAP EN ferisis cheat dee eine nag Melton: eae Ll aldo 1.086 56 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 Although an official procedure has been adopted for total sulphur determinations, the technique of the method as carried out by A. J. Patten, of Michigan, is so simple and satisfactory that his letter of trans- mittal is here given for the benefit of the association: For comparison, results are given on linseed meal by the sodium peroxid method, which I consider superior to the Schreiber method in that it is simpler and does not introduce so many factors that may disturb the precipitation of barium sulphate. The sodium peroxid method is very simple, and when carried out in the manner described should give no trouble. Sodium peroxid (official) method—Weigh the substance into a nickel crucible of 150 or 200 cc. capacity, add 10 ec. of water, and stir; then place the crucible in a pan of cold water and add slowly 5 grams of sodium peroxid stirring with a plati- num rod. (If the crucible is kept cool no difficulty is experienced in adding the peroxid.) Place the crucible on the hot plate and allow to remain until the con- tents are perfectly dry; again add 5 grams of sodium peroxid, and place the crucible covered tightly over the free flame and bring the contents to complete fusion. For the first 5 minutes it is well to have the flame just touching the bottom of the crucible, after which the full flame may be turned on. Sodium peroxid free from sulphur may be obtained without difficulty. Results reported by Mr. Schreiber in Circular 56 of the Bureau of Chemistry are in almost every case slightly higher by the peroxid method than by the proposed method, and, moreover, the peroxid method may be conducted with very much less attention to details than may the Schreiber method. In view of the excellent results secured by twelve analysts upon the present official method and reported to the association at its 1910 meet- ing, at which time the method was officially adopted, the referee would state that he heartily concurs in the opinion of Mr. Patten, and since no object seems to be attained by duplication of methods very similar, he would recommend that no further work be done on the Schreiber method. CALCIUM OXID AND MAGNESIUM OXID DETERMINATIONS. The favorable results of the preceding referee and of the present referee on the proposed extension of the official method for iron and alum- inum to include the determination of calcium and magnesium, where the usual minute quantities of manganese are present were responsible for the further work on this extension during the past year. It was intended to study any interference of large occurrences of manganese. The cal- cium was precipitated from the large volume resulting from the washing of iron and aluminum precipitate before elimination of manganese. It was found necessary to follow this procedure because the concentration of the solution under either acid or ammoniacal conditions resulted in the precipitation of calcium molybdate. By the usual redissolving and reprecipitation of the calcium oxalate, minute quantities of manganese were occluded from the abnormal occurrence of this element. Though the occluded manganomanganic oxid is sufficient to color the precipitate, 1915] McINTIRE AND CURRY: INORGANIC PLANT CONSTITUENTS 57 it was found as an average of 6 determinations by the referee to amount to only 0.4 mg. O. B. Winter, of Michigan, tested the calcium oxid precipitate and found 0.3 mg. of manganomanganic oxid, and J. P. Aumer, of Illinois, found 0.0019 mg. of manganomanganic oxid as an average of 8 determinations. It was the sense of the subcommittee passing upon 1912 recommendations that the minute occurrences (and oftentimes absence) of manganese would not vitiate any results within the most exacting margin of analytical error. If the normal occurrences of man- ganese were occluded in the same proportion as the abnormal amounts, by percentage, the occluded oxid could not be determined. The method used by Mr. Winter for determining the trace of manganomanganic oxid was not given. Mr. Aumer determined the oxid by direct and indirect removal of manganese with ammonium persulphate and by bromin pre- cipitation, while the referee dissolved the calcium oxid with very dilute nitric acid and weighed the remaining manganomanganic oxid. The procedure outlined is as follows: EXTENSION OF OFFICIAL MOLYBDIC METHOD FOR IRON AND ALUMINUM TO INCLUDE DETERMINATION OF MANGANESE, CALCIUM, AND MAGNESIUM. Keep the filtrate and washings from the first and second precipitations of iron and aluminum hydroxids to avolumeof 500 ce.; add ammonia and ammonium oxalate, redissolve the precipitate and reprecipitate as in cal- cium in soils (Bul. 107, Rev., p. 15). Then acidify the combined filtrates and washings from calcium determination and evaporate them to crys- tallization in a porcelain casserole. Place the casserole under a hot plate, protected from falling particles, and expel the ammonium salts. Take up with a few cubic centimeters of hydrochloric acid and water and filter off molybdic acid, washing the precipitate until it is free from chlorin. Bring the filtrate to volume of 100 cc., and add 1 or 2 drops of concentrated bromin. Make alkaline with ammonia and permit to stand several minutes without agitation. Filter off the precipitated manganese without boiling, wash, ignite, and weigh the precipitate as manganoman- ganic oxid. Concentrate the alkaline filtrate from the manganese separa- tion to 75 ee. and precipitate magnesium as in soils (Bul. 107, Rev., p. 16). Then dissolve this precipitate in hydrochloric acid and reprecipitate as above, filter, ignite and weigh as magnesium pyrophosphate. 58 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMists [Vol. J, No. 1 RESULTS OF COLLABORATION. Calcium oxid and magnesium oxid recovered from aliquots of 50 cc. of synthetic plant ash solution. ANALYST (27 DETERMINATIONS) | (17 DETERMINATIONS) grams per cent grams per cent O. B. Winter, East Lansing, Mich........... 0.0560 0.0463 0.0554 0.0459 0.0558 0.0457 Q20564.5 lease OR05607 ts eeeeerey OKO559' Foy Lie eee PASTE TAR Clay t sic.c stevoale ate ot Fekete erator ERENCE Sees 0.0559 0.0460 Maximum differences... on eeieee cet 0.0010 0.0006 Gaby Boltz. Wooster |Ohioy- asec eee On0547 ie ti easeee OF0550 10 le ate || ee ee OROS5O Fe ol) Bee eee QLO551 = PT |i Mase AVETA BER EA icloc eee eee rie tiee asiee kis tiene OLOS5ON" yr) 7 se eeten Maximumidifrerencess-eemen ashe tetera tne OFOOOS Ty ye Fal easacet: S. R. Mitchell, State College, N. M..........} 0.0595 0.0480 0.0593 0.0476 (OAK So oes AVETR DC es toists csc soseless acts Joveye cele aie rates shove abt terotel ae 0.0599 0.0478 Maximum'difierence.: 224.205. o2oese ese 0.0017 0.0004 doles dln Wid orien WON Ress seocoocn owe mwsat 0.0539 0.0451 0.0547 0.0475 0.0567 0.0490 0.0563 0.0477 0.0581 0.0485 0.0550 0.0486 0.0566 0.0459 0.0560 0.0466 IAN CLAP OS -2) Scie Ste Aton aos ce ies 10.0559 0.0474 Maximum difference..............-..+----- 0.0042 0.0039 W. H. McIntire, Knoxville, Tenn............ 0.0556 0.0454 0.0567 0.0451 0.0568 0.0454 0.0576 0.0443 OX0545.9 at 6 ia) heroes ORO55S 2S geese IAW CT AGES, eens fos isimiarae craven eRe oe crete ie 10.0563 10 0451 Maximumidifierencesnerenseee area eee 0.0031 0.0011 Grand aversgete «emcee roe enerne 0.0566 11.32 0.0466 9.32 Amount presenti iajaee ee oe ce ere ee 0.0569 11.38 0.0465 9.30 ETON ys nae See ee ee Eee —0.0003 —0.06 | +0.0001+0.02 1CaO purified. COMMENTS BY ANALYSTS. O. B. Winter: With proper care the method seems quite accurate. E. Van Alstine, for J. P. Aumer: I may say that by following the directions as given, when the calcium oxid was weighed it was colored by the presence of man- ganese. After being weighed, it was redissolved, the manganese removed, and the calcium again precipitated and weighed as calcium oxid, it being white this time. 1915) AVERITT: INSECTICIDES 59 S. R. Mitchell: Too much risk of loss of material by volatilization with ammonia fumes, by spurting out. Believe that some way should be suggested to avoid accumulation of so much ammonium salt. CONCLUSIONS. The results secured by the codperators are remarkably close, differing from theory by 0.38 mg. as an average of 27 determinations for cal- cium oxid and 0.1 mg. in the case of magnesium oxid as an average of 17 determinations. In view of the excellent results secured by the method during four years of study, and the slight occlusion of manganese even when its occurrence is made many times that of normal, the referee recommends the adoption of the method as provisional, with a view to its being made official. It is suggested, in conclusion, that if considered essential, purification of the calcium oxid precipitate be accomplished by solution of the lime by very dilute nitric acid, which does not attack the manganomanganic oxid. In this way occluded manganomanganic oxid may be easily and rapidly determined and correction made therefore. At 12.45 the convention adjourned until 2 p.m. MONDAY—AFTERNOON SESSION. At the opening of the afternoon session the president appointed the following committees: Committee to invite the Secretary of Agriculture and the Assistant Secre- tary to address the association: W. A. Withers, of North Carolina; L. L. Van Slyke, of New York, and F. T. Shutt, of Canada. Committee on nominations: R. J. Davidson, of Virginia; J. W. Kellogg, of Pennsylvania; and C. 8. Cathcart, of New Jersey. Committee on resolutions: J. M. Bartlett, of Maine; A. W. Blair, of New Jersey; and KE. W. Magruder, of Virginia. Auditing committee: J. P. Street, of Connecticut; H. D. Haskins, of Massachusetts; and W. H. McIntire, of Tennessee. REPORT ON INSECTICIDES. By 8S. D. Averitt, Referee. The work on insecticides for the past year, in accordance with the recommendations of the association has included a comparison of the methods for the analysis of lime-sulphur solutions proposed by the referee for 1911 with those proposed by the present referee in 1912, and a com- parison of the method of digestion of lead arsenate for water-soluble ar- senic proposed by the present referee, with the provisional 10 days’ 60 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMists [Vol. I, No. 1 digestion. The work on lime-sulphur solutions has required so much time that the second proposition has received very little attention. LIME-SULPHUR SOLUTIONS. For the comparison of the methods for the analysis of lime-sulphur solutions, two samples of known composition were made and sent out about the middle of February, 1913, to fifteen chemists who agreed to assist in this work. Two solutions, A and B, were used in making the samples of lime sulphur. A was a lime-sulphur solution made by boiling for 45 minutes in an atmosphere of hydrogen 100 grams of Kahlbaum’s calcium sulphid (CaS), 150 grams of precipitated sulphur, and one liter of distilled water. The calcium sulphid was tested for other metals and none were found. B was a solution of calcium thiosulphate which had no trace of sulphid. These solutions were very carefully analyzed, duplicates run in all cases, and in the case of A, both the iodin method and the zine chlorid method were employed in the determination of thiosulphate sulphur in order that there might be no question raised as to the accuracy of that determination and both methods gave the same figure. The following table gives the _results of that work: Solution A. serivinzueccw, |||) MORGSEEEEE® | peessuumets 7 (aut aa per cent per cent per cent per cent per cent 12.22 2.33 0.59 0.09 4.69 12.21 2.35 0.59 0.09 4.71 Average, 12.22 2.34 0.59 0.09 4.70 Thiosulphates by zine chlorid method: 0.57, 0.60, and 0.59 per cent; average, 0.59 per cent. Solution B. THIOSULPHATE SULPHATE 2) TOTAL SULERUR SULPHUR SULPHUR Determined Calculated per cent per cent per cent Ber cent per cent 5.42 5.37 0.08 4.82 4.85 5.46 5.37 0.08 4.84") ||) eee Average, 5.44 5.37 0.08 A838 - | sees As stated before, B was entirely free of sulphids. It was tested with both nickel sulphate and sodium nitroprussid. Sample 1 was made by adding to exactly 500 grams of A, exactly 500 grams of freshly-boiled and cooled distilled water. Sample 2 was made by adding to exactly 500 grams of A exactly 500 grams of B. These samples were kept bottled several weeks before transferring to the sample bottles sent to the workers. 1915] AVERITT: INSECTICIDES 61 Assuming that no change had taken place in the solutions, Sample 1 should have had the following composition: Total sulphur 6.11 per cent, thiosulphate sulphur 0.30 per cent, monosulphur equivalent 1.17 per cent, sulphate sulphur 0.05 per cent, lime (CaO) 2.35 per cent, sulphid sulphur (by difference) 5.77 per cent. Sample 2 should have had the following composition: Total sulphur 8.83 per cent, thiosulphate sulphur 2.99 per cent, monosulphur equivalent 1.17 per cent, sulphate sulphur 0.09 per cent, lime (CaQ) 4.77 per cent, sulphid sulphur 5.77 per cent. INSTRUCTIONS FOR COOPERATIVE WORK. Dear Sir: I am sending you by parcel post 2 samples of lime-sulphur solutions made carefully by myself; their composition is definitely known. The referee for 1911 holds that the methods proposed last year are not accurate and that those proposed by him in 1911 are so, and the association instructed that the methods be compared. The methods proposed in 1912 are those worked out by J. E. Harris at the Michigan Agricultural Experiment Station and published in Technical Bulletin No. 6 of that Station, with some changes in detail and manipu- lation which were found to facilitate the work. The methods of 1911 (Bur. Chem. Bul. 152, p. 70) may be called the zine chlorid methods, and those proposed last year (Bur. Chem. Bul. 162, pp. 36-37) the iodin methods, for distinction. My objections to the zine chlorid methods are based on the 1911 figures and on other work done by myself. The method for total sulphur is too indefinite and in the hands of an inexperienced worker may lead to bad results; the method for sul- phid sulphur involves the filtering and washing of zine sulphid, is long and tedious, and did not give accurate results on the 1911 samples; the method for thiosulphate sulphur did not give accurate results on the 1911 samples; the figures for sulphate sulphur in 1911 are not correct. The former referee denies that these objections are valid. The work of a half dozen good chemists this year, however, should settle the question. I believe the zine chlorid methods will give better results this year than in 1911 because the samples submitted this year offer better conditions for good results. I have found sodium peroxid a better oxidizing agent than hydrogen peroxid; it does the work in a fraction of the time and can be obtained sulphur free., I desire to call attention to the determination of lime (CaO). Many of the results last year were not correct. I regretted it very much as the lime affords the only check upon the analysis of a straight lime-sulphur solution as ordinarily made. The samples submitted this year are of such strength that 10 grams should be made to 100 ce. and 10 ce. aliquots used for the determinations in the iodin methods. Freshly-boiled and cooled distilled water should be used for this work in both methods. Do not fail, after having done the work, to give me your opinion of the methods. An analyst’s comments are often of nearly as much value to the referee as his figures, in that they show the advantages and disadvantages of a method which figures alone do not reveal. I think no change will take place in the samples, but it will be best to do the work as soon as you can find time for it. They should be protected from the air as much as possible during the work. Report results as soon as work is done. Your work will be very much appreciated. RESULTS OF COLLABORATION Reports have been coming in from 13 of the chemists assisting in this work since about March 1. Comparison of the iodin and zinc chlorid methods for the analysis of lime sulphur solutions. Sample 1. SULPHATE |, & Z Es E as) TOTAL SULPHID ee AND ge | eee S > SULPHUR | SULPHUR | SUUPBA‘ pues 3 E] (Cao) = S rs a Ree ee [ree eee | ee (eed le ee SSS ANALYST 2 2 z a |e z z z 3 5 = & “— o>) 3 ise] De & 3 & = = ‘e 2 2 5 eo a a a g a | az aS: | er Se Nh ist WS i ete a om ||) erik estan geting | oie wk | UE Bed 2 = ° I ° ds 2 By ag ofS} mo] 9 s2, = 2s! 3 fal BTU a oe an |i SS | Se RS aE IS se |e we |] 2 per | per | per | per | per | per | per | per | per | per | per | per cent | cent | cent | cent | cent | cent | cent | cent | cent | cent | cent | cent E. W. Gaither, Wooster, Ohio | 6.20 | 6.12 | 5.85| 5.45| 0.38] 0.32] 0.05 1.21 | 2.32 | 2.32] 5.76 6.12 | 6.01 | 5.78} 15.19] 0.32] 0.26] 0.05 1.19 | 2.3i | 22.46 | 25.72 6.15 | 6.08 | 5.82 |15.28] 0.35) 0.30] 0.04] 0.06] 1.20 | 2.34 ao 6.15 | 6.10 | 5.81 | 45.26] 0.35] 0.32] 0.04 1.20 | 2.32 & 6.13 | 6.08 | 5.81 | 15.37] 0.35) 0.32 Bee 1.20 | 2.32 0. B. Winter, va, N.Y.| 6.15 | 6.19 | 5.98 inter, Geneve 6.15 | 6.16 | 5.86 6.14 5.94 G. P. Gray, Berkeley, Calif. o > oe > Be aw yeaa) Com nn a > a a es . rm D. M. Nelson, Chicago, Il... rrr Cl + Soe A. J. Patten, and W. C. Marti, East Lansing, Michisoeccses- cece eee S._D. Averitt, Lexington, Ky It H. H. Hill, R. J. Davidson, W. B. Ellett, H. O. Till- man, Blacksburg, Va....| 6.10 | 6.15 0.26 | 0.39} 10.01] 0.03] 1.15 | 2.52 | 2.33] 5.89 6.18 | 6.23 0.26 | 0.39 | 10.01 1.15 | 2.54 | 22.48 | 25.81 6.12 | 6.24 Oo eee 1.15 | 2.52 ARS 6.22 | 6.25 0229) eee 1.22 | 2.52 WEG | ose 0.26) .... 1.21 980 6.24 ous General average Theoretical. ............. aed | his tt 1 Not included in the general average. 2? Calculations using the zine chlorid figures for thiosulphate sulphur and sulphate sulphur. 62 1915] AVERITT: INSECTICIDES 63 Sample 2. 5 & THIO- Re EE IE Ba TOTAL a TOTAL BUNBHID | Werner AND £ B=] oie iu SULPHUR | SULPHUR SULPHITE |0 4 > SULPHUR Aone NE 5 5 (CaO) A gi 8 eee = ANALYST ‘ 4 Be eas eee = = = Z| |@a g# 18 3s a) iS a 2 2 | 2 a 4 os 5 ey) 12h aS] RS | a = A s A & q iB Bp | Be Com?) ol?) = ° = =) = 2 = Sl 2 ee) oo |] moO is=) & so) a iso) a as) 2 a Be n i) & i) 4 S 4 S = [a Ola per | per | per | per | per | per | per | per | per | per | per | per cent | cent | cent | cent | cent | cent | cent | cent | cent | cent | cent | cent E. W. Gaither, Wooster, COL NTOVEE Sener RENO Ran 8.90 | 8.81 | 5.95 | 5.42] 3.07] 2.94] 0.08] 0.08] 1.18 | 4.82 | 4.85] 5.74 8.84 | 8.78 | 5.65] 15.09] 3.01] 2.82] 0.07] 0.08] 1.18 | 4.84 | 24.67 | 25.82 8.87 | 8.70 | 5.81] 15.24] 3.06] 2.86] 0.08] 0.04] 1.18 | 4.83 oe 8.86 | 8.66 | 5.80] 15.26 | 3.07] 2.88] 0.06] 0.05 | 1.18 | 4.83 8.84] .... | 5.86/15.36] 3.07] 2.85 aus 1.18 O. B. Winter, Geneva, N. Y.| 8.83 | 8.78 | 5.98 | 5.60] 3.00} 2.98] 0.05] 0.07 | 1.22 | 4.77 | 4.83] 5.79 8.86 | 8.83 | 5.94] 5.53] 2.98] 2.96] 0.06] 0.12} 1.22 | 4.79 | 24.92 | 25.73 8.84 5.76| 5.74] 2.99] 2.99 1,22 | 4.77 acc 2.99 S06 1.22 ||... G. P. Gray, Berkeley, Calit. ee 8.45 | 5.86] 5.82} 3.02 ae 0.07 D. M. Nelson, Chicago, Ill. . 8.72 | 8.70 | 16.33] 5.70] 3.02] 2.77] 0.15] 0.10} 1.15 | 4.70 | 4.89] 5.54 8.70 | 8.74 |15.19] 5.69] 3.02] 2.84] 0.13] 0.05{ 1.15 | 4.70 | 24.60 | 25.84 8.69 | 8.76 | 14.31] 5.65] 3.06] 2.83] 0.13] ....] 1.13 ( 4.72] .... 5 5.67 2.79 dine cisée. alatere 5.75 S00 aera A.J. Pattenand W. C. Marti, East Lansing, Mich...... 8.81 | 8.79 | 5.85] 5.83] 2.94] 12.58] 0.07] 0.05] 1.15 | 4.70} 4.70] 5.80 H. H. Hill, R. J. Davidson, W. B. Ellett, H. O. Till- man, Blacksburg, Va..... 8.71 | 8.91 3.01} 3.09) 0.12] 0.10] 1.14 | 4.84 | 4.87] 5.67 8.91 | 8.85 3.07 | 12.56 | 0.11] 0.11] 1.14 | 4.83 | 24.88 | 25.97 8.88 | 8.92 3.07] .. ORD Peas, (eed eese aaa EX Al| cso 0.12 1.17 | 5.12 siti ac thos Eales 1.17 ne 1.12 W. J. Morgan, Washington, De Cre eosesonsenesccene 8.81 | 8.90 | 5.70 | 16.06 | 13.12 | 12.63 | 0.11] 0.051 1.16 | 4.61 | 4.961 5.53 8.84 | 8.94 | 5.43 | 16.17 | 13.20 | 12.63 | 0.10] 0.06 | 1.13 | 4.63 | 24.39 | 26.23 557 13.23 1.13 | 4.68 nt} |\tooue 13,23 1.14 | 4.70 Joos Sct) Hees Msn R. C. Roark, Washington, Bi Crsacees Mean eae Nee -00 5.72 13.14 | 12.48 | 0.0 Ste 5.90 13.20 | 12.54} 0.0 5.84 3.05} ....] 0.0 5.87 13.11 0.1 5.78 ae 0.0! oor 0.1 General average............ 8.81 | 8.76 | 5.82] 5.77] 3.02) 2.89] 0.09] 0.07] 1.17 | 4.75 | 4.86] 5.69 sisheoretical../s:ssmesecciaven SIRE eee) Daal ctsce || 2009) mcs \OCOGI| Seme, I) edn ang 7/aeae 6102505 1 Not included in the general average. ? Calculations using the zine chlorid figures for thiosulphate sulphur and sulphate sulphur. 64 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMIsTs [Vol. I, No. 1 COMMENTS OF ANALYSTS. E. W. Gaither: Since I have not had any extended experience with the analysis of insecticides, I do not feel free to criticize the two methods very extensively. I will say, however, that judging from my results, the iodin method seems the more accurate. From the standpoint of ease and speed of manipulation it is preferable. O. B. Winter: (1) Sodium peroxid as used in the iodin method seems to be the better oxidizing agent. Its use saves time, and it can be obtained sulphur-free more easily than can hydrogen peroxid. (2) There should be no difficulty in getting the end point when titrating with iodin for either the monosulphid or thiosulphate sulphur in the iodin method, and this method is much more convenient than the zine chlorid method. (3) The iodin method for estimating sulphate and sulphite sulphur gives agreeing results, while the zinc chlorid method does not. In Sample1the results agree in both methods and with each other but in Sample 2 the zine chlorid method gives widely-varying results. When the solution is heated the results are likely to run high. (4) The check on the work by calculating the lime (CaO) that is com- bined with the different forms of sulphur is a great advantage. (5) In estimating the sulphid sulphur, the iodin method does away withthe device for removing clear test portions in order to get the end point when adding ammoniacal zine chlorid. Getting this end point is tedious. Filtering and washing the free sulphur in the iodin method is done much more easily than filtering and washing the zine sulphid. (Here I find it advantageous to allow the free sulphur to stand in the solution for several hours, stirring it occasionally to remove from the sides and bottom of the containing vessel. This overcomes the stickiness that sometimes exists, and it filters clear without difficulty.) In general, I believe the iodin method for analyz- ing lime-sulphur solutions is more accurate than the zine chlorid method, and it is more convenient. G. P. Gray: (1) Total sulphur.—From the experience in this laboratory on these and other samples, equally accurate results are obtained by either method. The method of oxidation with sodium peroxid seems decidedly preferable on account of greater rapidity and the elimination of the blank. (2) Sulphid sulphur.—Little preference is shown in either of the methods where the direct determination is made. (3) Jodin titration —The results seem to give somewhat high figures in this work, but it is probable that this could be rectified with more practice in determin- ing the end point. Its simplicity and rapidity certainly commend its use. Suggest taking larger aliquots for titration. A.V.H.Mory: We found good points and bad points in both the hydrogen peroxid and the sodium peroxid methods for total sulphur. While the sodium peroxid method is shorter, still we were unable to get sodium peroxid from a well-known local supply house which did not contain about one-half as much sulphur as the ordinary commercial hydrogen peroxid obtained without reference to its sulphur content, and the measuring of an exact quantity for the purpose of running a blank was a rather difficult matter. The sodium peroxid was represented to be the best obtainable and contained about 0.03 per cent of sulphur. While our experience with the iodin method for sulphid sulphur was not at all reassuring, we were led to entertain a different sort of opinion of the iodin method in the case of the thiosulphate sulphur and the sulphate sulphur, the manipulation being decidedly rapid and convenient and conducive to concordant results. It is to be noted, however, that the results obtained by the iodin method were higher than by the zine chlorid method ineach case. We do not feel qualified to venture an opinion as to which set of results is nearer the truth. There was little or no difficulty experienced in arriving at an end point in the titration for the monosulphur equivalent. 1915] AVERITT: INSECTICIDES 65 A.J. Patten: I very strongly favor the Harris methods. They are simpler and much easier of manipulation. R. J. Davidson: Now, as to any criticism of the iodin method as compared with the official method, I do not feel that I have ever done enough with the method to be in a position to criticize it as it should be done. I think the titration method is certainly a very rapid one, and probably where one has had some experience there will be no difficulty in determining the end point. This, I think, is the chief diffi- culty that there is with the method, because every excess used here in the mono- sulphid diminishes the amount of the thiosulphate. H.H. Hill: We have gone over these methods quite thoroughly and it seems as though the iodin method is superior both in manipulation and accuracy to those proposed by the former referree. Sodium peroxid as an oxidizing agent is much more advantageous, as it can be obtained free from sulphur, while with the hydrogen peroxid much time was required in the preliminary treatment with barium carbonate. Calculating the lime is also another advantage, as it checks the determined lime. With the new method the time of operation is shortened considerably. R. C. Roark, associate referee: In commenting on these results, I wish first to criticise the samples sent. A’ good commercial concentrated lime-sulphur solution will contain about 25 per cent of total sulphur, hence your samples do not represent the commercial product, and methods applicable to dilute solutions will not neces- sarily be suitable for concentrated solutions. The samples sent out were so dilute that they decomposed on standing a few days and solutions made up from them for analysis at the rate of 10 grams to 100 ce. decomposed almost immediately. The monosulphur equivalent may be determined with fair accuracy if the titration is done in a not too dilute solution, but the monosulphur equivalent of itself is a de- termination of no value. Any error in the monosulphur equivalent titration affects the thiosulphate titration. Furthermore, it affects the amount of sulphur precipi- tated and hence the sulphid sulphur determination; and there is reason to believe that an excessive amount of iodin may cause an increased formation of sulphates, hence affecting the sulphate sulphur determination. Thus an error in the first titration will make an error in every subsequent determination made on the same aliquot. Results for the thiosulphate sulphur as determined by the Michigan method are of no value due to the titration of hydrosulphid and hydroxyhydrosul- phid sulphur as thiosulphate sulphur, as shown by C. C. McDonnell at last year’s meeting of the association. Relative to the claim that the lime calculated from the iodin titration agrees with the lime determined, I will say that our analyses of these samples, as well as your analysis of the samples I sent you last May, show that this claim is not supported by the facts. Lime calculated does not agree with lime determined. DISCUSSION. Before discussing the results, the referee wishes to call attention to the fact that all but one of the codperating chemists, in commenting upon the methods, prefer the iodin methods because of the ease and rapidity with which the work is done. One analyst criticizes the samples, regardless of the fact that they favor the methods which he champions. The word- ing and tone of his criticism, however, convict him of carelessness in handling the samples. He asserts that the monosulphur equivalent of itself is a determination of no value. In the absence of proof to this effect, the use the other chemists have made of it seems to prove the contrary. 66 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. J, No. 1 His comments in regard to thiosulphate sulphur by the iodin titration method lose much of their force in the face of his figures by the zine chlorid method, and the figures of the other analysts do not sustain his comments. His comment in regard to the calculated and determined lime is not confirmed by the figures. While the writer is advocating the iodin methods he wants it understood that this is not a personal matter with him, and that he is interested only to the extent that the association will adopt accurate, workable methods. The elimination of some results was in order that neither method should be done an injustice. As for the results themselves, it should be noted that they show con- clusively that dilution has not changed the composition as the average results agree almost perfectly with the theoretical values in all cases ex- cept total sulphur and thiosulphate sulphur in Sample 2 by the zinc chlorid methods, and these do not vary much. In all but a few cases in both samples the calculated lime agrees well with the determined lime, the general average of both methods being close to the theoretical in Sample 1 and the iodin figure in Sample 2. The general average of the calculated lime by the zine chlorid methods in Sample 2 is very low, which fact is explained readily since these values are calculated on each analyst’s average, and two or three bad figures in the same direction influence the average. The general average of the sulphid sulphur by difference, by the iodin methods, agrees closely with the theoretical in Sample 1, and fairly well in Sample 2; by the zine chlorid methods, agrees fairly well in Sample 1 and not well in Sample 2. The referee has been absolutely fair to both methods in the omission of bad results from the average. It must be noted, however, that the general averages are largely in favor of the iodin methods, notwithstand- ing the fact that these samples favor the zine chlorid methods, and the great majority of the chemists get good results by those methods. REFEREE’S COMMENT ON 1912 PAPER BY McDONNELL. In a paper entitled “‘Composition and Methods of Analysis of Lime- sulphur Solutions,” read before the last meeting of the association, in opposition to the iodin method, the former referee on insecticides made some statements that subsequent work in my laboratory has shown to be untenable when the experiments upon which he bases his statements are carried on under reasonable conditions. “The cause of the inaccuracy of the direct iodin titration method” for thiosulphate sulphur he bases upon the work done by Divers and Schimidzu (J. Chem. Soc., 1884, 45: 270-291). I shall prove to the satis- faction of any disinterested chemist that the statements in this connection are not in accord with the facts as shown by work in my laboratory. (1) It is claimed that checking the analysis by the determined lime and 1915] AVERITT: INSECTICIDES 67 the lime calculated from the monosulphur equivalent, the thiosulphate sul- phur, and the sulphate and sulphite sulphur, is erroneous because dilution decreases the monosulphur equivalent value and increases the thiosul- phate value and that unless the dilution is great the two errors counter- balance each other. Under moderate dilution the monosulphur equiva- lent is not decreased as I shall show later; neither is the thiosulphate increased. It was the fact that tenth-normal zine chlorid and tenth-nor- mail iodin gave the same figures for monosulphur equivalent that led Mr. Harris to investigate the direct titration with iodin. Tartar and Bradley of the Oregon Agricultural Experiment Station have shown that standard hydrochloric acid with Methyl Orange as indicator, will give the same value for monosulphur equivalent as standard iodin or standard zine chlorid. I have corroborated the work of both Harris, and Tartar and Bradley. There is no question about these facts, and if standard zine chlorid gives the correct sulphid sulphur, standard iodin does also. Under the head of ‘Effects of dilution on monosulphur equivalent and thiosulphate” the former referee makes dilutions of 20, 60 and 250 ce. in one case, and 50, 100, and 200 cc. in the other, and for the 250 ce. and 200 ce. dilutions he does have a decreased monosulphur equivalent and an increased thiosulphate figure. Under these dilutions decomposition took place before the titrations could be made. Moreover, in such ab- normal dilutions the end points are not distinct. This is self evident when the solution is in the process of breaking down. Here are the facts in regard to the effect of dilution. From two lime- sulphur solutions, one rather weak and the other much stronger, called for convenience No. 1 and No. 2, I took 20 grams of No. 1 and made it to 100 grams with carbon-dioxid-free water, a dilution of 1 to 4. Of the diluted solution, I took 5 grams, equivalent to 1 gram of original solution and titrated it with twentieth-normal iodin with the following results: DILUTIONS IODIN FOR MONOSULPHUR EQUIVALENT ce. cc. IODIN FOR THIOSULPHATE SULPHUR cc. 30 12.00 0.75 60 12.00 0.75 100 11.80 0.85 The same treatment of 20 grams of No. 2 gave the following results: IODIN FOR MONOSULPHUR IODIN FOR THIOSULPHATE DILUTIONS EQUIVALENT SULPHUR cc. cc. cc. 30 16.30 6.50 60 16.30 6.60 100 16.40 6.50 68 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMIsTs [Vol. J, No. 1 These titrations give in No. 1, 0.96, 0.96 and 0.94 per cent of mono- sulphur equivalent, and 0.24, 0.24 and 0.27 per cent of thiosulphate; in No. 2, 1.30, 1.80 and 1.31 per cent of monosulphur equivalent, and 2.08, 2.11 and 2.08 per cent of thiosulphate. It was found that when the dilution amounted to 100 ce. it was neces- sary to titrate rapidly and use an external indicator for the first end point and starch solution for the other. With dilutions of 150 to 200 ce. the decomposition began so quickly and was so rapid that any values might be obtained depending upon the rate of titration. When a dilution of 150 to 200 cc. was made with filtered tap water the sulphur would begin to come out by the time it could all be added in 50 ec. portions. In the case of ordinary distilled water there was a short interval between the addition of the water and the precipitation of sul- phur. With freshly-boiled and cooled distilled water there was a sensibly longer interval between the addition of the water and the precipitation of sulphur. This indicates that the dilution is not the only factor involved in the decomposition. From the above figures it appears that any reasonable dilution gives the same values for monosulphur equivalent and thiosulphate sulphur. At the time the above work was being done (February 18, 1913), 20 grams of these samples were made to 100 grams with carbon dioxid free water, bottled, sealed, and left in the laboratory; No. 1 until June 18, 1913, four months later, when the monosulphur equivalent and thiosulphate sulphur were determined by the iodin titration with the following results: Monosulphur equivalent 0.20, 0.20 and 0.20 per cent; thiosulphate sulphur 0.054, 0.054, 0.044 per cent; average monosulphur equivalent 0.20 per cent, average thiosulphate 0.05 per cent, which in the original is equiva- lent to 1 per cent of monosulphur equivalent and 0.20 per cent of thio- sulphate sulphur which is in perfect agreement with what was found in the original four months before. No. 2 remained in the laboratory until October 13, 1913, eight months after it was made, when the monosulphur equivalent and thiosulphate sulphur were determined as in No. 1 with the following results: Mono- sulphur equivalent 0.262 and 0.264 per cent; thiosulphate sulphur 0.41 and 0.41 per cent; average monosulphur equivalent 0.263 per cent; aver- age thiosulphate sulphur 0.41 per cent, which in the original is equivalent to 1.32 per cent of monosulphur equivalent and 2.05 per cent of thiosulphate sulphur—a difference of 0.02 per cent in the monosulphur equivalent and 0.04 per cent in the thiosulphate sulphur from the findings in the original eight months before. As stated before, the dilutions made on the original lime-sulphur solu- tion in order to make the present association samples did not affect the composition, as the work of a large majority of the chemists shows that the 1915] AVERITT: INSECTICIDES 69 composition of Samples 1 and 2 agrees well with the theoretical values, assuming that dilution made no change in the composition. No further arguments are needed to show that moderate dilution has no effect upon the monosulphur equivalent and thiosulphate content of a lime-sulphur solution. The dilutions made by the former referee are entirely abnormal, and were not contemplated in the instructions for this work. (2) It is stated upon the authority of Divers and Schimidzu, as noted above, that the inaccuracy of the iodin titrations is due to the fact that there is present in the diluted solution not only calcium polysulphid, calcium thiosulphate, calcium sulphate and sulphite, but calcium hy- droxid, calcium hydrosulphid (Ca(SH)2), calcium hydroxyhydrosulphid (Ca(SH)(OH)), and hydrogen sulphid (H.S). Four equations repre- senting the decomposition are given as follows: (1) CaS;-+2H,0 =Ca(OH).+48-+H.8 (2) Ca(OH)2+2H2S = Ca(SH)2+2H20 (3) Ca(SH)2+H,0 =Ca(SH) (OH) +H.8 (4) Ca(SH) (OH) +H.0 =Ca(OH)2+H.8 It is stated, however, that “‘These decompositions proceed only when the hydrogen sulphid produced can escape or become diluted.’’ When the first step in this decomposition takes place, which is represented by equation (1), free sulphur must be precipitated. This does not occur unless oxidation takes place which is shown as follows: Many and large dilutions have been made in this laboratory of lime-sulphur solutions with freshly-boiled and cooled distilled water, bottled and sealed at once and allowed to stand in the laboratory for weeks and months and even years, and are as clear in the end as when first made. There are few solutions more stable than lime-sulphur under any reasonable dilution, provided it is protected from the air or from oxidation in other ways. To test this decomposition further, two portions of 10 cc. each of a lime-sulphur solution were diluted with 30 ce. of distilled water in a small narrow neck flask and a strip of moistened lead acetate paper sus- pended above the liquid, the flask stoppered and allowed to stand. At the end of two hours no indication of hydrogen sulphid was visible, at the end of three hours a faint trace was observed, and at the end of four hours it was distinct but the paper was not black, showing that the decom- position proceeds very slowly. It is stated that the diluted lime sulphur has in addition to the caleium polysulphid and thiosulphate originally present, hydrogen sulphid, cal- cium hydrosulphid and caleium hydroxyhydrosulphid, all of which form clear solutions and when titrated with iodin the yellow color due to the polysulphids disappears before these colorless sulphur compounds have 70 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 been attacked and that they are titrated and calculated as thiosulphate, thus giving a low monosulphur equivalent and a high thiosulphate value. This statement is absurd in the face of the facts shown above. More- over, if as stated by the former referee, zine chlorid precipitates the sulphur of these colorless compounds of sulphur (hydrosulphid and hy- droxyhydrosulphid) and iodin does not, the average chemist will have difficulty in explaining the fact that standard iodin and standard zine chlorid give the same values for monosulphur equivalent. The referee has titrated sodium sulphid (a clear colorless solution) without the use of any indicator at all, and has been able to tell within the limits of experimental error, how much sulphid and thiosulphate sul- phur were present in the solution. He has had others mix standard so- dium sulphid and standard thiosulphate and upon titrating the solution has determined both very accurately. This is not at all difficult and may be done by any careful analyst. The first addition of iodin will produce a yellow color due to the formation of polysulphid. When this disappears upon further addition of iodin the sulphid sulphur titration is ended. The appearance of the yellow color due to the excess of iodin shows the end of the thiosulphate titration. As a positive proof that there is no calcium hydrosulphid (Ca(SH)o), or calcium hydroxyhydrosulphid (Ca(SH)(OH) ) in the solution after the monosulphur equivalent titration with iodin is complete, the writer has repeatedly tested the solution with nickel sulphate and found no sulphids present. (3) It is stated that the end points in the iodin titrations are not defi- nite, the error in the monosulphur equivalent titration amounting to 0.1 ec. of tenth-normal iodin, and that for the thiosulphate titration being another 0.1 cc., or a total error of 0.2 ec. of tenth-normal iodin, which is asserted by the critic to be a fair degree of accuracy for this titration. Such errors in the end points will not occur if the directions are followed. An error of 0.05 ec. might be made in the monosulphur equivalent titration as no end point is more definite than that for the thiosulphate titration. The results of a large majority of the workers this year show that this degree of accuracy is not only possible, but that it is very easy to attain. With a little experience the end points are just as definite as that of stand- ard alkali against standard acid with Methyl Orange or phenolphthalein as indicator. The referee does not claim, however, this definiteness of end points in a lime-sulphur solution so diluted that decomposition begins before the titrations can be made, which will invariably follow dilutions of 150 to 200 cc. as made by the former referee. (4) In regard to the rapidity of the methods, it is stated that it re- quires very little more time to make a determination of thiosulphate 1915] AVERITT: INSECTICIDES 71 sulphur by the zine chlorid methods than by the direct titration with iodin. To this I will reply that I have made three direct titrations in the time required to make one by the zine chlorid method. The sulphid sulphur determination by the iodin method requires no more time than by the zine chlorid and is much easier of manipulation. The comments of the analysts who have aided in the work this year afford a very definite answer to this objection to the iodin methods. REFEREE’S RESULTS ON COMMERCIAL LIME-SULPHUR SOLUTIONS. During the past year the referee has had occasion to work upon a half dozen or more samples of commercial concentrates representing the product of four or five of the leading manufacturers of lime sulphur in the country. These samples have been investigated mainly with reference to two points, first, the thiosulphate sulphur as determined by the iodin methods, by the zine chlorid methods as approved by the association in 1911, and by a modification of the zine chlorid method as approved; second, the lime as calculated from the iodin titrations and the sulphate and sulphite sulphur compared with the determined lime. The following facts were found to be true without exception: The thio- sulphate sulphur by the zine chlorid method as approved was always from 0.30 to 0.50 per cent lower than by the iodin method. When, how- ever, 2.5 grams instead of 10 grams were made to 100 e.c. the zine chlorid method gave results which compared very favorably with the iodin method; when 5 grams were made to 100 ec. the results were intermediate between the figures by the method as approved and the iodin figures. The simplest conclusion to be drawn from these facts is that the thiosulphate is held up by the bulky precipitate of zine sulphid. As to the second point investigated, it was found that the lime as calculated from the iodin titrations and the sulphate and sulphite sulphur agrees very closely with the determined lime. About six weeks ago the referee received from a Kentucky company a sample of their lime-sulphur solution which they are making and selling. They are using a very pure lime and a relatively pure water for this work and the result is a typical straight lime-sulphur solution, a sample of which was sent to the associate referee with the request that he determine by the approved zine chlorid methods the total sulphur, thiosulphate sulphur, sulphate and sulphite sulphur, and lime. The referee worked this sample by both the iodin and approved zine chlorid methods and also the thio- sulphate sulphur by the zine chlorid method with 2.5 grams to 100 ce. instead of 10 grams to 100 cc. This work was done in order that the comparison of the two methods might be made complete and that the two points brought out by the referee’s work on concentrates as given above might be brought before the association in figures and discussed. The iz ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 writer feels that this is the proper and only way to settle the present con- tention definitely and permanently. Following is a tabulation of the results: SULPHID 4 : “ uimE (CaO) se nono THIOSULPHATE SULPHATE AND 5 : SULPHUR \SULPHITE SULPHUR TOTAL DIFFERENCE | SULPHUR Calculated SULPHUR EQUIVA- u ee 2 ss ——— Se Deters | ae ENT mined . Zine - Zine . Zine : Zine . n - - = Todin | chlorid Teds ehlorid | 1042 | chlorid Todin | chlorid | per cent | per cent) per cent) per cent per cent per cent per cent | per cent) per cent| per cent| per cent 25.80 | 24.31] 24.72) 5.03 1.33 1.04 0.17 | 0.03 | 10.33} 10.27] 9.77 5.03 1.33 1.04 0.16 | 0.04 | 10.32 1.04 2.5 grams to 100 ce. | 1.30 | | 1.30 Sulphid sulphur: iodin method, 25.80—(1.33-+0.17) =24.30 per cent; zine chlorid method, 25.80—(1.04+0.03) = 24.73 per cent. Lime (CaQ) by calculation, iodin methods =5.03 X1.75+1.83 X0.87+-0.17 X1.75 or 8.81+1.16+0.30=10.27 per cent; difference between determined and calculated lime, 0.05 per cent. Lime (CaO) by calculation, zine chlorid methods =5.03 X1.75+1.04 X0.875+0.03 or 8.81+0.91+0.05=9.77 per cent; difference between determined and calculated lime, 0.55 per cent. From the above table it will be seen that the calculated lime by the iodin method is in close agreement with the determined lime, there being a difference of only 0.05 per cent. On the other hand the lime calculated, using the zine chlorid figures for thiosulphate sulphur and sulphate sulphur, and the iodin figures for monosulphur equivalent (standard hydrochloric acid and standard zine chlorid give the same value as standard iodin for monosulphur equivalent), is 0.55 per cent below the determined value. Now it is perfectly well understood that the calcium and not the sulphur is reacting with the iodin; therefore, the iodin titration is a measure of the calcium combined as polysulphids, and the lime (CaO) equivalent of the calcium (Ca) in the polysulphids plus the lime (CaQ) equivalent of the calcium in the thiosulphate, plus the lime (CaO) equivalent of the cal- cium in the sulphate and sulphite must of necessity be equal to the lime equivalent of all the calcium in the solution. And since in a straight lime-sulphur solution the lime is all in combination as the polysulphids, thiosulphate, sulphate, and sulphite, it follows that the determined and calculated lime must be equal. In the approved zinc chlorid methods the thiosulphate and sulphite figures must be erroneous, because they do not give back the determined lime. The iodin titrations must be right unless - 1915] AVERITT: INSECTICIDES 73 we suppose calcium to exist in a lime-sulphur solution in other combina- tions than those here considered, and as to this the chemical reactions involved in the making of a lime-sulphur solution preclude the forma- tion of any other compounds. It has been conclusively shown that no change occurs in the solution so long as it is not allowed to oxidize, and any reasonable dilution with carbon dioxid free water produces no change unless oxidation occurs. The inevitable conclusion, therefore, is that the thiosulphate and sulphate sulphur results in a concentrate as determined by the approved zine chlorid methods, are erroneous. There is another way to show the inaccuracy of the zine chlorid methods and the accuracy of the iodin methods, as follows: The sulphid sulphur by the zine chlorid methods divided by the monosulphur equivalent = 24.73+5.03=4.92, indicating that about 92 per cent of the sulphid sulphur is in the form of calcium pentasulphid. Now, if the sulphur equivalent of the calcium in the polysulphid as shown by the zine chlorid methods is divided into the sulphid sulphur, 24.73+5.38=4.60, indicat- ing that only 60 per cent of the sulphid sulphur is in the form of calcium pentasulphid. The first ratio indicates an excellent concentrate, the second a rather poor one. Both conditions can not be true, and it is known that this sample is a good concentrate. On the other hand, the sulphid sulphur, as shown by the iodin methods divided by the monosulphur equivalent, 24.30+5.03 =4.83, indicating a good concentrate. The sulphur equivalent of the calcium in the poly- sulphid as shown by the iodin methods, divided into the sulphid sulphur, 24.30+5.06=4.80, which is entirely consistent with the facts. As stated above, this is a typical lime-sulphur solution and the facts brought out in this discussion were found to be true for all the other samples of con- centrates worked. Last May, following a request for a sample of a commercial concen- trate that had been recently worked by the zine chlorid methods in the Insecticide and Fungicide Laboratory of the Bureau of Chemistry, I received two samples of lime-sulphur solution marked “A” and “B” from the Department of Agriculture. I made the analysis of these samples by the iodin methods as I had proposed to the associate referee for the purpose of comparing the results by the two methods. On account of an error in my first determinations of lime I was led to think they were straight commercial concentrates as my erroneous determinations of lime checked the calculated lime. When the figures by the two methods were compared it was found that on Sample A the zine chlorid methods gave 0.50 per cent less lime than I found by both calculation and determination by the iodin methods, and 0.50 per cent less thiosulphate sulphur was found by the zine chlorid than by the iodin methods. The following table shows the average of all results by both methods: 74 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. J, No. 1 THIOSULPHATE SULPHUR SULPHATE SULPHUR LIME (CaO) MONOSULPHUR EQUIVALENT r s : 2 é Todin Zine chlorid Todin Zine chlorid | Determined | Calculated per cent per cent per cent per cent per cent per cent per cent 5.02 1.50 1.00 0.22 0.04 9.98 10.48 When I found that there was only 9.98 per cent of lime in this sample I knew it could not be a straight lime sulphur. The iodin titrations were repeated and found to agree perfectly with those first made. Now the question arose as to what metal was reacting with the iodin; 2.5 grams were made to 100 ce. and the thiosulphate sulphur determined by the zine chlorid method and 1.48 per cent of thiosulphate sulphur found, so that there was approximately 1.50 per cent of thiosulphate sulphur in the solution. A determination of alkalies was made and a pure sodium salt to the extent of 1.06 per cent of sodium chlorid was found in the solution. It was thought best to test the solution for chlorids, which was done, but only a very small amount, equivalent to 0.03 per cent of sodium chlorid was found, so that it was evident that it was a sodium salt reacting with the iodin and that it was the thiosulphate was very probable from other circumstances. The sodium in the sodium chlorid was calculated to sul- phur in thiosulphate and found to be equivalent to 0.56 per cent of sulphur, which left 0.94 per cent of sulphur in the form of calcium thiosulphate. Using this value for the thiosulphate sulphur instead of 1.50 per cent, the calculated lime is 9.99 per cent. It seems to the referee, in view of these facts, that it might be said with just as much propriety that a volumetric determination and a gravi- metric determination of lime will not check as to say that the calculated lime and determined lime will not check in a straight lime-sulphur solution. WATER-SOLUBLE ARSENIC IN LEAD ARSENATE. In regard to the recommendation relative to the time and temperature of digestion for water-soluble arsenic in lead arsenate, the referee, after consultation with the associate referee decided that the recommendations as made in this report would meet the approval of all concerned. It is to be noted that the methods of analysis are not in question at all. The time of digestion is pretty well established and a few degrees in tempera- ture is not very material in view of the fact that both time of digestion and temperature are entirely arbitrary. The referee is well aware and has pointed out the fact that what we are determining and ealling soluble arsenic is largely soluble lead arsenate, but as both time and temperature of digestion are not absolute factors, it is probably a matter of small moment. 1915] AVERITT: INSECTICIDES 75 RECOMMENDATIONS. Tt is reeommended— (1) That the method for total sulphur in lime-sulphur solutions as given in Bureau of Chemistry Circular 108, page 2 (a) be changed in the fifth line (p. 3, line 2) to read “2 grams of sodium peroxid,’’ and as changed, be made an official method. Hydrogen peroxid may be used as the oxidizing agent as follows: Add to the aliquot of solution 3 cc. of saturated solution (1: 1) sodium hydroxid followed by 50 cc. of hydrogen peroxid and let stand on steam or water bath one-half hour, then acidify with hydrochloric acid and complete the determination as usual. (In case hydrogen peroxid and sodium hydrate are used, a blank must be run for the sulphur contained in them.) (2) That the method for thiosulphate sulphur given in Bureau of Chemistry Circular 108, page 3 (b) be made an official method, the fol- lowing note to be added: ‘‘In the case of a concentrate twentieth-normal iodin, instead of tenth-normal, may be used to advantage, and the exact end point for the monosulphur equivalent then determined as follows: Near the appearance of the yellow color take up a drop on a small glass rod and apply to a few drops of nickel sulphate solution on a porcelain plate.” (3) That the method for sulphate and sulphite sulphur as given in Bureau of Chemistry Circular 108, page 3 (c), be made an official method. (4) That the method for sulphid sulphur as given in Bureau of Chem- istry Circular 108, page 3 (d), have the following words in the third line, ‘and dissolve the sulphur in 15 ce. (1:3) sodium hydrate,” replaced with the following words: “flask with about 5 ec. of water and add 15 ee. (1:3) sodium hydroxid;” and in the fifth line strike out “‘1 to 14 hours” and re- place it with “3% to 1 hour,” and as thus changed be made an official method. (5) That the method for total lime (CaO) in solution as given in Bureau of Chemistry Circular 108, page 3 (e), have the last line continued as follows: “or determine volumetrically with tenth-normal potassium permanganate,” and as thus changed be made an official method. The sulphur may be precipitated with hydrochloric acid, the solution boiled or heated on water or steam bath until the hydrogen sulphid is driven off, the sulphur filtered off and washed, and the lime determined in the filtrate as above. (6) That the following method of digestion for water-soluble arsenic in lead arsenate be made a provisional method, and that the present pro- — visional method of digestion be dropped: Water-soluble arsenic—Weigh to 0.01 gram about 4 grams of paste; place in a tightly-stoppered flask or bottle with 250 cc. of freshly-boiled and cooled distilled water per gram and keep at 32°C. for 24 hours, shaking 76 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 well every hour during the working day (8 times in all), filtering at the end of 24 hours. Use 250 ce. of the clear filtrate for the determination; add 0.5 ec. of sulphuric acid and proceed as directed under water-soluble arsenic oxid, Bulletin 107, Revised, page 240. It is important that the solution shall be perfectly clear and the titrations carefully made. Make corrections for iodin necessary to produce the same color, using same chemicals and volumes. (7) That the method by C. C. Hedges of Cornell University (J. Ind. Eng. Chem., 1909, 1: 208), for the determination of arsenious oxid in Paris green and other insecticides, be compared by the next referee with the official method now in use. (8) That the Lloyd method for nicotin in tobacco and tobacco extracts be compared with the official method (Kisslings). Lloyd method—Weigh into a strong 200 ce. beaker 4 to 5 grams of a 3 to 7 per cent extract or 0.3 to 1 gram of a 40 per cent extract, add 2 to 3 ec. of water and sufficient ferric hydroxid mixture (equal weights ferric hydroxid and sodium acid carbonate) both ammonia-free, to make a thick paste; add 12 ee. of petrolic ether or washed gasoline, stir the whole with any convenient instrument for 2 or 3 minutes, allow to settle, and decant the solution into a separatory funnel. Repeat the extraction 4 or 5 times, decreasing the amount of ether 1 ce. each time. To the contents of the separatory funnel add 25 ec. of tenth-normal sulphuric acid and 25 ce. of water; shake the whole strongly and allow to settle. Draw off the acid solution into a deep porcelain dish, wash the contents of the funnel once or twice with 25 cc. of water and add the washings to the dish, the contents of which is titrated with tenth-normal sodium hydroxid, using neutral litmus or cochineal as indicator. 1 ce N/10 H,.SO,=0.0162 nicotin. A COMPARISON OF THE IODIN TITRATION AND ZINC CHLORID METHODS FOR THE ANALYSIS OF LIME- SULPHUR SOLUTIONS. By R. C. Roark, Associate Referee. For the determination of the various forms of sulphur and lime in lime-sulphur solutions, different methods have been proposed, of which two, namely, the iodin titration method as worked out by the Michigan Agricultural Experiment Station,! and by the referee on insecticides,? and the zine chlorid method of Sutton’ as modified by Haywood‘! and by McDonnell’ are now before this association for consideration. 1 Mich. Agr. Exper. Sta. Tech. Bul. 6. ? Bur. Chem. Bul. 162, pp. 29-30. 3 Volumetric Analysis, 10th ed., pp. 342-345. 4 J. Amer. Chem. Soc., 1905, 27: 244. 5 Bur. Chem. Bul. 152, p. 70; Bul. 162, p. 49. 1915] ROARK: ANALYSIS OF LIME-SULPHUR SOLUTIONS 77 The two methods briefly are as follows: In the iodin titration method, a portion of the solution suitably diluted is titrated with standard iodin solution, the first reaction being, theoretically, CaS,+I1. => Cal.+8,. This titration determines the monosulphur equivalent, and from this value the amount of lime combined with sulphur as polysulphids may be calculated. This titration does not determine the sulphur present in the solution as sulphid or as polysulphid; it is really a titration of the lime combined as calcium polysulphid. After the end point of this titration is reached, which is determined by the disappearance of the yellow color of the polysulphids, the titration is continued until the solution turns yel- low from excess of iodin. By this second titration the thiosulphate sulphur is determined according to the reaction: 2CaS.03+ L => CalI.+ CaS.0.z, the iodin acting as its own indicator. Sulphid sulphur is determined by filtering off the sulphur precipitated by the first iodin titration, oxidizing it to sulphate with sodium peroxid and determining as barium sulphate in the usual way. Sulphate and sulphite sulphur (which has been oxi- dized to sulphate sulphur by the iodin) are determined in the filtrate by precipitation with barium chlorid. Lime is determined on a separate aliquot of the solution, the sulphur being oxidized to sulphate by sodium peroxid and lime being precipitated from the solution of calcium sulphate by ammonium oxalate in the usual way. In the zine chlorid method, thiosulphate sulphur is determined by precipitating sulphids with an ammoniacal solution of zine chlorid, filter- ing, neutralizing the filtrate with hydrochloric acid and titrating with standard iodin solution with starch as an indicator. Sulphid sulphur is determined by precipitating as above; the precipitate of zine sulphid is then washed, oxidized to sulphate by means of sodium or hydrogen peroxid, and the sulphur weighed as barium sulphate. Sulphate sulphur is deter- mined in the filtrate from the thiosulphate titration by precipitation with barium chlorid. (Sulphite sulphur, if present in the original solution, is oxidized to sulphate by the iodin solution and would, therefore, affect both the thiosulphate and sulphate determinations. If it is present at all, however, it is only in traces and may be ignored.) Lime is determined by treating a portion of the solution with hydrochloric acid, filtering from sulphur and precipitating with ammonium oxalate from the solution of its chlorid. It will be noticed that with the iodin titration method, all the deter- minations with the exception of lime are made on the same aliquot, whereas with the zine chlorid method, every determination is made on a separate 78 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMIsTs [Vol. J, No. 1 aliquot with the exception of thiosulphate and sulphate sulphur. The determination of total sulphur according to the two methods differs only in the use of the oxidizing agent, which is sodium peroxid in the iodin titration method and hydrogen peroxid in the zine chlorid method.1_ Both give accurate results and so are not discussed in this paper. OBJECTIONS TO THE IODIN TITRATION METHOD. Assuming in the iodin titration method that the forms of sulphur pres- ent include only the polysulphid, the thiosulphate, the sulphite, and the sulphate of calcium (an assumption that later we shall show to be incor- rect) the method is open to the following objections: (1) In the titration for the monosulphur equivalent and thiosulphate sulphur, dilution of the aliquot being titrated has a marked effect on the results, as has been pointed out by McDonnell (See Bur. Chem. Bul. 162, p. 40). The present referee has found that dilution has an effect on titration as shown in a letter to C. C. McDonnell under date of February 21, 1913: I found that when the dilution amounted to 100 cc. it was necessary to titrate rapidly and use an external indicator for the first end point and starch solution for the other. With dilutions of 150 to 200 cc. I found that the decomposition began so quickly and was so rapid that any values might be obtained, depending upon the rate of titration. It has been urged that such high dilutions were not contemplated by the method, but a dilution of a 10 ce. aliquot to 60 ce. or to 100 ce. can not be considered an excessive dilution, yet in one case a dilution from 20 ee. to 60 ce. decreased the monosulphur equivalent from 6.00 to 5.94 and increased the percentage of thiosulphate sulphur from 1.73 to 2.00; in another case a dilution from 50 ce. to 100 ec. decreased the monosulphur equivalent from 4.59 to 4.40 and increased the thiosulphate sulphur from 1.34 to 1.75. (2) In the second place, the end point of either of these titrations can not, at the best, be determined closer than 0.10 ce. of tenth-normal iodin solution each, which on an aliquot representing 0.25 gram of sample (5 grams to 200 cc., a dilution recommended by the referee for concen- trates), would cause an error of 0.07 for the monosulphur equivalent value and an error of 0.28 per cent for the thiosulphate sulphur value. It may be that these errors are compensating, but no method is to be considered accurate that depends on an uncertain compensation of errors. Unless the end point of the monosulphur equivalent titration be de- termined by means of nickel sulphate as an outside indicator, an error of more than 0.10 ec. of tenth-normal iodin may easily be made. If an outside indicator is used, the withdrawal of portions of the solution (more 1 Avery, Bur. Chem. Bul. 90, p. 105. 1915] ROARK: ANALYSIS OF LIME-SULPHUR SOLUTIONS 79 especially when it is of small volume as it must be to avoid the effect of di- lution) introduces an error in the titration for thiosulphate sulphur and also in the determination of sulphid sulphur, as particles of suspended sulphur are removed with the solution when a portion is withdrawn for testing. Furthermore, the titration of thiosulphate sulphur with iodin as its own indicator, especially in the presence of precipitated sulphur, can not be as accurate as where starch paste is used to show the end point. But if starch paste is used, it interferes with the filtration of the precipitated sulphur so that this is another instance of what occurs throughout the iodin titration method, that is, the use of a less accurate method for one determination so that the succeeding determination made on the same aliquot will not be spoiled. That the end points of these iodin titrations are difficult of determina- tion is shown by the results of the different analysts codperating with the referee in 1912 (See Bur. Chem. Bul. 162, pp. 33-34). These results varied so that out of fourteen values for thiosulphate sulphur, the referee discarded seven, or one-half. Results for monosulphur equivalent on sample No. 2 by two analysts were also thrown out. (83) In the determination of sulphid sulphur by the iodin titration method, the sulphur precipitated by the iodin titration is filtered off, oxidized with sodium peroxid, and precipitated as barium sulphate. This sulphur can not be filtered at once as it will pass through the finest filter paper. It has been found necessary to allow the sulphur to stand for sev- eral hours, or overnight, before filtering, thus causing a delay in the completion of the analysis—a delay that is not met with in the zinc chlorid method where the precipitate of zine sulphid can be filtered immediately. The filtration of sulphur at best is an unsatisfactory process and the chances of error through manipulation as compared to the filtration of zine sulphid, are relatively as 3 to 1, which is the ratio of the molecular weight of zine sulphid to the atomic weight of sulphur. (4) Results for sulphate sulphur are higher by the iodin titration than by the zine chlorid method, as is shown in the following table: METHOD SAMPLE Iodin titration Zine chlorid LOTS PAM O PAL CAIN On ena. stess 0.07 0.03 ; 0.09 0.02 HOTS VAS OAR CANO 2th cere. 0.11 0.05 0.10 0.06 IMIS CINO} 138265 rycen ecacks ists 0.16 0.03 0.16 0.03 Mis Cre NOT 14799 Naeem 0.22 0.04 0.21 0.08 80 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 As an explanation of this, it would seem probable that in the iodin titration some of the sulphur other than sulphite sulphur is oxidized to sulphate. It is not unreasonable to suppose that such a strong oxidizing agent as iodin does oxidize some of the sulphur compounds present to sulphate. (5) For the determination of lime, the referee recommends using an aliquot which will give in the case of a concentrate containing approxi- mately 10 per cent calcium oxid, only 0.025 gram of ignited lime. It is maintained by the writer that for accurate results a sufficiently large aliquot should be taken to give at least 0.100 gram of ignited lime. No objection, theoretical or otherwise, can be brought against this. OBJECTIONS TO THE ZINC CHLORID METHOD. (1) It has been objected that in the precipitation of sulphids by the ammoniacal zine chlorid solution, the zine sulphid precipitate may carry down and hold some of the thiosulphate, thus making the results for sul- phid sulphur high and the results for thiosulphate sulphur low. Both calcium and zine thiosulphates are extremely soluble in water and are stable unless the solution is boiled, whch is not done in this method, so there is no reason for their not appearing in the filtrate from the zine sulphid precipitate due to insolubility or to decomposition. Moreover, McDonnell! has found that on adding thiosulphate sulphur as sodium thiosulphate to lime sulphur solutions, up to 50 times the amount occurring in the sample, the whole amount added could be recovered when deter- mined according to this method. This proves that no thiosulphate sul- phur is destroyed or held up by the zine sulphid precipitate. Further proof of this statement is to be found in the results of the analysts on the 1911 lime-sulphur samples.2 Sample No. 2 was the same as No. 1, ex- cept for the addition of 2.02 per cent of thiosulphate sulphur as sodium thiosulphate. The average difference in thiosulphate sulphur in these two samples from 19 determinations is 1.98 per cent, which agrees closely with the theoretical 2.02 per cent. (2) As regards tediousness, zine sulphid may be filtered at once and with accuracy, whereas in the iodin titration method the precipitated sulphur must stand sometime before being filtered, and then with the probability of losing appreciable quantities. RESUME OF OBJECTIONS. To sum up the criticisms of the probable errors in the two methods due to manipulation. 1 Bur. Chem. Bul. 162, p. 41. ? Bur. Chem. Bul. 152, p. 71. 1915] ROARK: ANALYSIS OF LIME-SULPHUR SOLUTIONS 81 (1) In the titration of thiosulphate sulphur by the iodin titration method, it is shown that errors may arise, (a) from reading the end point of the monosulphur equivalent titration more especially when nickel sulphate is not used as an outside indicator and any error in this titration causes a considerable error in the titration for thiosulphate sulphur which follows (1 ee.n/10 [=0.0016 gram of sulphid sulphur, but =0.0064 gram of thiosulphate sulphur): (b) from reading the end point of the thiosulphate titration where starch is not used as an indicator; (c) from the effect of even moderate dilution of the aliquot which decreases the values for monosulphur equivalent and increases those for thiosulphate sulphur. In the zine chlorid method only one iodin titration is made in deter- mining thiosulphate sulphur and starch is used as an indicator, conditions more favorable to accuracy than are those of the iodin titration method. (2) That in the determination of sulphid sulphur, zine sulphid may be filtered at once, whereas sulphur precipitated by iodin must stand for sev- eral hours before filtering, and even then with the chance of loss by running through the paper. Moreover, zine sulphid may be washed free of thio- sulphate, whereas it is yet to be shown that in filtering precipitated sulphur which has stood in a solution containing sulphur compounds, that the sulphur particles do not hold some of these sulphur compounds. (3) In the zine chlorid method, sulphate sulphur is determined in the filtrate from the thiosulphate sulphur titration where only a few cubic centimeters of iodin have been used, whereas in the iodin titration method sulphate sulphur is determined after the solution containing unstable and easily oxidizable sulphur compounds has been exposed to the action of a large amount of iodin, conditions favorable to the oxidation of sulphur, in forms additional to sulphite, to sulphate. (4) According to the method of the referee, lime is precipitated from a solution in which it is present as sulphate, whereas according to the method of the associate referee, it is precipitated from a solution of the chlorid and an aliquot is used which gives about four times as much ignited lime as will be obtained according to the referee’s method. COMPOSITION OF LIME-SULPHUR SOLUTIONS. In discussing the relative merits of the iodin titration and of the zine chlorid methods for the analysis of lime-sulphur solutions, it has been assumed that the only calcium sulphur compounds present were polysul- phids, thiosulphate, and small amounts of sulphite and sulphate. On carefully looking up the literature, however, it is found that such is un- doubtedly not the case. Review briefly what compounds have been shown to be present in pure lime-sulphur solutions, and what effect these compounds, other than those already mentioned, would have on the determinations as made according 82 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 to the two methods. Beginning with the sulphids, the simplest is caleium monosulphid (CaS), which can not exist in solution, however, as it is de- composed by water as follows: 2CaS-++2H,0 = Ca(SH)2-+Ca(OH)s! The next in series is calcium bisulphid (CaS.). Tartar and Bradley? obtained a compound whose composition corresponded very closely to this formula by evaporating lime-sulphur solution in an atmosphere of hydrogen over sulphuric acid in the dark and at a low temperature, washing the residue with carbon disulphid until all free sulphur was extracted and drying in a current of hydrogen. While calcium trisulphid (CaS;) has not been isolated from a lime- sulphur solution, it seems probable that it would be formed by the boil- ing together of lime and sulphur. The very recent work of Tartar? shows that potassium trisulphid (K.S3) is formed in the primary reaction of sulphur with potassium hydroxid in heated aqueous solution, and from analogy, calcium trisulphid should be formed in the preparation of a lime- sulphur solution. Calcium tetrasulphid (CaS,) is formed by boiling together 1 equivalent of calcium monosulphid with 3 equivalents of sulphur,’ or by the action of sulphur on lime in heated aqueous solution.® Calcium pentasulphid (CaS;) may be formed by dissolving sulphur in a solution of calcium hydrosulphid,*® by boiling calcium monosulphid with sulphur,’ or by boiling sulphur with lime.* No higher sulphid of calcium than the pentasulphid is known to exist. In this connection it may be well to recall that the principal objection brought against the zine chlorid method last year by the present referee was that, according to analyses by this method a sulphid of calcium higher than the pentasulphid was shown to exist. This claim was based on the results on the two association samples for 1911, in which ratios of 5.3 to 1 and 5.6 to 1 were found for sulphid sulphur and lime.’ These ratios, however, were obtained by dividing the values for sulphid sulphur, determined according to the zine chlorid method, by the values for mono- sulphur equivalent determined by the iodin titration method. A slight . error in this latter value (which could have been made from the effect of dilution alone, as the determination was made before the work of Mc- ‘ Abegg’s Handbuch der anorg. chemie, 1905, 2 (2): 116. > J. Ind. Eng. Chem., 1910, 2: 275. 3 J. Amer. Chem. Soc. 1913, 35: 1746. ‘Schéne, Pogg. Ann., 1862, 117: 75. ® Yartar and Bradley, loc. cit., p. 274. 6 Divers and Schimidzu, J. Chem. Soc., 1884, 45: 284. ' Gmelin-Kraut, Handbuch der anorg. chemie, 1909, 2 (2): 222. 8 Tartar and Bradley, loc. cit., p. 277. 9 Letter to C. C. McDonnell under date of November 17, 1911. 1915] ROARK: ANALYSIS OF LIME-SULPHUR SOLUTIONS 83 Donnell showing the marked effect of dilution had been brought to the attention of the referee) would change the ratios materially. On association sample No. 1 for 1911 the present referee obtained a value for lime (CaO) of 6.05 per cent, agreeing well with the value calcu- lated from his results by the iodin titration method, namely, 6.04 per cent. According to the closely agreeing analyses by one of the Bureau of Chemistry chemists, however, the lime in this sample was only 5.53 per cent, or 0.52 per cent less than found by the present referee. From this it is seen that the original objection of the present referee to the zine chlorid method was based on faulty premises. McDonnell! showed that among the results of analyses of 29 lime- sulphur solutions tabulated in Michigan Technical Bulletin No. 6, there were 5 having ratios of total sulphid sulphur to monosulphur equivalent greater than 5 to 1, the highest being 5.11 to 1. On the other hand, from the analyses of 33 samples of lime-sulphur solutions examined by the zine chlorid method, only 1 showed a ratio of polysulphid sulphur to lime ereater than 5 to 1, namely 5.08 to 1. The work that was done by Van Slyke? in showing that no sulphid higher than the penta is present in lime-sulphur solution, was done ac- cording to the zine chlorid method, and if this method does not give correct results, this very conclusion would be erroneous. In addition to the sulphids of calcium, there are in a lime-sulphur solu- tion certain oxy-sulphur compounds, of which calcium thiosulphate is found in largest amount. First prepared by Herschell® in 1819, it is formed whenever lime and sulphur are brought together in heated aqueous solution, the amount formed being dependent on concentration and length of boiling.4 Calcium sulphite (CaSO3) exists in small amount in lime-sulphur solution as shown by Van Slyke.® Calcium sulphate (CaSO,) exists in traces only in lime-sulphur solution.® Hydrogen sulphid is present in small amount in nearly all commercial concentrated lime-sulphur solutions. In the preparation of lime-sulphur solution, hydrogen sulphid is given off in large quantities,’ so that it is natural to expect its presence in the finished product. Furthermore, on dilution of a lime-sulphur solution some hydrogen sulphid is produced according to the reaction: Sree CeO Hae tee 1 Bur. of Chem. Bul. 162, p. 38. 2N. Y. Agr. Exper. Sta. (Goaey a) Bul. 319, p. 394. 3 Riebirch Phil. J., 1819, 1: 8. ‘ Fordos and Gelis, ‘Annales de Chimie et de Physique 1846, (3) 18: 86; Senderens, Bul. de la Société Chimique de Paris, 1891 (3) 6: 800; Tartar and Bradley, loc. cit. 5 N.Y. Agr. Exper. Sta. (Geneva) Bul. 319, p. 383. 6 Van Slyke, loc. cit., pp. 383 and 396. 7 Herschell, Fordos and Gelis, Senderens, Schéne, loc. cit., and Firard, Compt. Rend., 1863, 56: 797. 84 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 In addition to these well-known compounds, there are others which are formed by the interaction of these, or by the action of water and oxygen. There is some hydrosulphid formed by the action of hydrogen sulphid with lime: Ca(OH):+2H2S = Ca(SH)2+ 2H20! Also, hydrosulphid is formed as an intermediate product in the reaction between polysulphid and water: 2CaS,+ 2H:O = Ca(SH)2 a Ca(OH)2.+28,-1. The presence of hydrosulphid in lime-sulphur solution has been disputed, and in support of this contention the work of Tartar and Bradley is often cited. A careful reading of their article, however, shows that they do not make the statement that no hydrosulphid is present. They state: “Our tests show the absence of appreciable quantities of hydrosulphid.’’ This conclusion is based on the results of one quantitative experiment. They determined sulphid sulphur by titration with standard ammoniacal zine chlorid solution and also by determining the hydrogen sulphid evolved upon addition of dilute acid. By the first method, only one-half the sulphur combined as hydrosulphid would be shown, as it would be pre- cipitated as zine hydrosulphid, but calculated as zine sulphid (ZnS); all the sulphid sulphur would be evolved as hydrogen sulphid upon addition of acid. In this way they found in a solution 3.42 grams of sulphur per 100 ec. by the sulphid sulphur method and 3.44 grams of sulphur per 100 ec. by the hydrogen sulphid method, or 0.020 gram more. Sulphur present as hydrosulphid would be twice 0.020 gram =0.040 gram, or 1.13 per cent of the total sulphid sulphur present, surely more than a trace. Moreover, on dilution, a condition which prevails in the course of analysis, another oxysulphur compound is formed, namely, hydroxyhydrosulphid: Ca(SH)2+H20 = Ca(SH)(OH) +H:S It is doubtful if more than a trace of this compound exists at any one time, but as these reactions are reversible it is formed as an intermediate product, (1) in the reaction between lime and hydrogen sulphid, and (2) in the decomposition of polysulphids by water. ° Consider what effect these hydrosulphur compounds would have on the iodin titration method. Hydrogen sulphid would be titrated as monosul- phur equivalent, making the results for that high, and, of course, the value of lime calculated from it high. Calcium hydrosulphid dissolves in water, 1 Divers and Schimidzu, J. Chem. Soc., 1884, 45: 271. 2 Loc. cit. p. 277. 3 Loe. cit. p. 274. 1915] ROARK: ANALYSIS OF LIME-SULPHUR SOLUTIONS 85 forming a colorless solution, hence by the iodin titration method it could escape titration with the polysulphids and be determined with, and cal- culated as, thiosulphate sulphur, which, as results show, is higher by the iodin titration than by the zine chlorid method. The existence of a hydro- sulphid is also indicated by the fact that in the monosulphur equivalent titration the end point is always reached quicker when taken by the dis- appearance of the yellow color than when nickel sulphate is used as an outside indicator. By the zine chlorid method any hydrosulphid or any hydroxyhydrosul- phid present would be precipitated as zine sulphid and would not appear, nor be titrated, in the filtrate as thiosulphate sulphur. Certain other oxysulphur compounds of calcium are known. Herschell! by boiling together for one hour, sulphur and lime in 20 parts of water, obtained orange-colored crystals which, after drying over sulphuric acid in a vacuum showed upon analysis the composition CaS.0.4H.O. As later determined by Schone? the formula is Cas8,03:12H.O, or, accord- ing to the analysis of Geuther* Ca;8;02 with either 10 or 11 molecules of water. In addition to this compound, which is known as Herschell’s crystals, there is another calcium sulphur complex which has been found in lime-sulphur solution, namely, Biichner’s crystals, whose composition is represented by the formula Cas8,0418H20* or by Ca;8;0314 or 15H.0.5 Tartar and Bradley, in preparing pure lime-sulphur solutions, observed the formation of orange-red needle-shaped crystals which they stated “were undoubtedly the oxysulphids of calcium.’’ That these crystals occurred in solutions such as might be met with at any time is shown by the analysis. One solution contained 5 grams of lime (CaO) and 10.5 grams of total sulphur; the other 9.45 grams of lime (CaO) and 20 grams of total sulphur per 100 ce. If these oxysulphids compounds exist in solutions such as these in so large amount as to crystal- lize out on cooling, may we not expect them in appreciable quantities in most every lime-sulphur solution? While not more than a trace of free calcium hydroxid (Ca(OH)s) may be present in a freshly-prepared lime- sulphur solution,’ it is formed in appreciable amounts upon dilution according to the reaction. CaS;+2H20 = Ca(OH)2+H2S+48 and would be present, therefore, in lime-sulphur solutions which had stood for some time and become partially decomposed. } Edinburgh Phil., J., 1819, 1: 8. 2 Pogg. Ann., 1862, 117: 78. 3 Ann. Chem. (Leibig), 1884, 224: 190. ‘Schone, loc, cit., p. 85. 5 Geuther, loc. cit., p. 193. 6 J. Ind. Eng. Chem., 1910, 2: p. 273. 7 Tartar and Bradley, loc. cit., p. 273. 86 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 As pointed out by McDonnell,! there is no reason why calcium hy- droxid can not be added to a lime-sulphur solution after its preparation and marketed in such form. McDonnell showed that by adding 15 ce. of clear lime water to a lime-sulphur solution having a monosulphur equivalent of 5.26 and 3.54 per cent of thiosulphur, the value for mono- sulphur equivalent was increased to 6.54 and thiosulphate to 4.12 per cent. Undoubtedly the lime in the complex oxysulphid compounds previously mentioned is loosely combined and on titration of the solution with iodin would consume it, and, hence, add to the values for thiosulphate. SUMMARY. In addition to the tetra and pentasulphid, the thiosulphate and the sul- phite and sulphate of calcium, admitted by the referee to be present in a concentrated lime-sulphur solution, there are some calcium bisulphid, very probably some calcium trisulphid, undoubtedly small amounts of hydrogen sulphid and calcium hydrosulphid, traces of free caleium hy- droxid, and appreciable amounts of oxysulphid compounds of variable composition. On dilution for analysis, there will be formed more hydro- sulphid and more free calcium hydroxid, together with small amounts of calcium hydroxyhydrosulphid. The referee claims that the accuracy of the iodin titration method is proved in that the lime (CaQ) in solution eal- culated from the values for monosulphur equivalent, thiosulphate and sulphate sulphur agrees with the lime determination. If, as the referee assumes, the solution contained only polysulphids, thiosulphate, sulphite and sulphate of calcium, this would be true and the two values for lime should agree. As we have just shown, however, a lime-sulphur solution contains in addition to these compounds small amounts of hydrogen sul- phid and calcium hydrosulphid, calcium bisulphid, calcium trisulphid, traces of free calcium hydroxid and appreciable quantities of complex oxysulphid compounds; therefore, even if the iodin titrations gave abso- lutely correct results for monosulphur equivalent and thiosulphate sulphur (which is never true, owing to the effect of dilution, error in reading end points, etc.) lime (CaO) could not be calculated because of the presence of these other sulphur compounds of calcium. The results of analysis of some commercial concentrated lime-sulphur solutions bear out what should be expected from this theory. Lime was determined according to the method recommended by the author, in which the calcium is precipitated from a solution of its chlorid, using a sufficiently large aliquot to give an accurately weighable ignited precipi- tate. Sulphate sulphur was determined according to the zine chlorid 1 Bur. Chem. Bul. 162, p. 41. 1916] ROARK: ANALYSIS OF LIME-SULPHUR SOLUTIONS 87 method. If determined according to the iodin titration method the re- sults would have been higher, thus making the results for calculated lime still further from those of determined lime. Thiosulphate sulphur and monosulphur equivalent were determined by the iodin titration method, nickel sulphate being used as an outside indicator for the first titration and the end point of the second titration being checked always by adding a little starch paste. Ten cubic centi- meter aliquots, representing about 0.4 gram of concentrated solution, were diluted to 35 ce. only before titration to avoid the effect of dilution. The results given are the mean of closely-agreeing determinations. Analyses of concentrated lime sulphur solutions. oe IODIN TITRATION ZINC CHLORID METHOD eG LIME (CaQ) LABORATORY Kee Total | Sulphid | Sulphate sulphur | sulphur | sulphur . . Monosul- Ene eaaiees phur Deter- | Caleu- | Differ- equiva- | mined lated ence sulphur | sulphur lant per cent | percent | per cent | per cent | per cent | per cent | per cent | per cent | per cent 13826 25.00 | 24.24} 0.03 | 0.87 44} 5.00] 9.75 | 10.05 | —0.30 14156 25.63 | 24.72 | 0.04] 0.76 .62 | 5.08 | 10.02 | 10.387 | —0.35 14799 24.51 | 23.47 | 0.04] 0.81 .09 | 9.96 | 10.29 | —0.33 14883 24.10 | 22.78 | 0.07] 0.94 -69 | 9.25 | 9.44! —0.19 14994 25.17 | 24.25 | 0.05 .68 | 10.17 | 10.24 | —0.07 15156 25.24 | 24.01} 0.04 90 | 10.09 | 10.38 | —0.29 15157 26.08 | 25.53 | 0.03 .20 | 10.18 | 10.58 | —0.40 15819 24.88 | 23.80 | 0.03 92 | 9.78 | 10.14 | —0.36 15825 21.85 | 20.92 | 0.04 50 | 9.01} 9.68 | —0.67 15874 24.99 | 24.34 | 0.04 : 9.95 | —0.22 16043 24.60 | 23.64] 0.04 78 | 9.68 | 10.02 | —0.34 16140 21.78 | 20.14] 0.03 13} 9.09} 9.36 | —0.27 16143 22.51 | 21.16 | 0.03 oo) |) 9/19) 975.) —0256 16146 22.97 | 22.14 | 0.04 Gye SElGhe9= 26) 5 —Os10 16235 25.38 | 24.29 | 0.03 .18 | 10.09 | 10.30 | —0.21 16244 24.73 | 23.82 | 0.03 99 | 9.77 | 10.03 | —0.26 16397 24.85 | 24.23 | 0.04 .99 | 9.75 | 10.25 | —0.50 SCOCOOCOCORRREORFRCOrHS S e one ne en ee no ~I for) oO I ~j In every case it is seen that the value for lime calculated from results obtained by the iodin titration method is higher than that determined gravimetrically, the difference ranging from 0.07 to 0.67 per cent with an average of 0.32 per cent. An explanation of this will be given later. In addition to these, the results of other workers show that the calcu- lated and determined values for lime do not agree. The following table gives the values for lime as determined by Tartar and Bradley! in the lime-sulphur solutions studied by them. The values for calculated lime have been figured from their results for sulphite and sulphate sulphur, thiosulphate and ‘“‘sulphid”’ sulphur, which they determined by titration with zine chlorid solution: 1J. Ind. Eng. Chem., 1910, 2: 273, 275, 276. 88 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. J, No. 1 Results by Tartar and Bradley. GRAMS PER 100 cc. LIME DESCRIPTION OF SAMPLE 3 co} © # 3 S e/2| 5 A|;|o;A (Gommercial Sse tmcncs catticies seiteiccyasieiweaetoeeys , URI GAC ASE TUR DOSSCORERCoD EB tas Ss abodes 2.12) 11.77|+0.35 Commercial............ 14.07/—1.21 Gommercial 35: ne. dec spor cn cin w setoeioateels wale arene ene wecleies eeaye ciate aicielasisleiny Gene oie cece eae 12.73)/—0.35 55 grams of lime + 110 grams of sulphur + 400 cc. water boiled 1} hours............... §.49| 9.47/+0.02 55 grams of lime + 55 grams of sulphur + 450 cc. water boiled 1 hour................ 4.99) 5.27|—0.22 60 grams of lime + 110 grams of sulphur + 450 cc. water boiled 1 hour................ 9.45) 9.65)—0.20 55 grams of lime + 110 grams of sulphur + 450 cc. water boiled 1 hour................ 9.68) 9.69/—0.01 60 grams of lime + 110 grams of sulphur + 450 cc. water boiled 2} hours............... 8.24) 8.50/—0.26 55 grams ot lime + 55 grams of sulphur + 430 ce. water boiled 1 hour................ 5.15) 5.34/—0.19 55 grams of lime + 55 grams of sulphur + 450 cc. water boiled 24 hours. .............. 5.10} 5.32|/—0.22 60 grams of lime + 110 grams of sulphur + 450 cc. water boiled 1 hour 9.45) 9.65)—0.20 60 grams of lime + 110 grams of sulphur + 450 cc. water boiled 2} hours 8.39} 8.49—0.10 Assuming that titration with zine chlorid solution yields the same re- sults for monosulphur equivalent, or “sulphid”’ sulphur, as it is termed by Tartar and Bradley, as when -iodin solution is used (an assumption held to be true by Harris and by Averitt), and assuming further, accord- ing to the contention of the referee, that only the polysulphid, the thio- sulphate, the sulphite and sulphate of calcium are present, it is seen from the above results that the zine chlorid method gives, if anything, high results for thiosulphate sulphur. From all the work that has been done on the comparison of the two methods, however, thiosulphate sulphur by the iodin titration method is invariably higher than thiosulphate sulphur by the zine chlorid method and so must necessarily be farther from the true figures. The official samples of lime-sulphur solution sent out this year, accord- ing to the analyses of the Bureau chemists, do not show agreement between the calculated and determined values for lime: LIME SAMPLE AND ANALYST Determined Calculated Difference per cent per cent per cent Sample 1 Analyst ie se ccc ose 2.24 2.56 —0.32 Analyst: 2coeceose be 2.23 2.53 —0.30 Sample 2 Analyst. a oeiccccec oe - 4.66 4.99 —0.33 LWTOIN 74-Scapono done 4.65 4.95 —0.30 1915] ROARK: ANALYSIS OF LIME-SULPHUR SOLUTIONS 89 The present referee himself, from the analysis of two commercial con- centrated lime-sulphur solutions as received from the manufacturers, found that lime determined did not agree with lime calculated: Analysis by Averitt. LIME SAMPLE Determined Calculated Difference per cent per cent per cent ‘SA”’ (Misc. No. 14799)... 9°99 10.49 —0.50 “B”’ (Mise. No. 13826).... 19.76 10.12 —0.36 1 The mean of ten closely-agreeing results by 3 analysts determined according to both the Bureau of Chemistry and the referee’s methods. In Sample ‘‘A”’ nine closely-agreeing results by the same analysts gave a mean of 9.97 per cent. The analyses presented in Michigan Technical Bulletin No. 6 do not bear out the contention of the author of the iodin titration method that lime may be accurately computed from the results for the various forms of sulphur. In the table on the following page are given the values for determined lime as they appear in the various tables in this bulletin, together with the values for calculated lime which have been figured from the values given for monosulphid sulphur, thiosulphate sulphur and sulphite and sulphate sulphur. All results have been figured to per — cent by weight. It will be noticed that the differences between the two values for lime vary from —1.18 to +0.44, a variation of 1.62. From this table it is seen that in twenty-seven out of the thirty-two samples, the difference between the calculated and determined values for lime is 0.05 per cent or more; in eighteen, the difference is 0.10 per cent or more; in fourteen the difference is 0.15 per cent or more; in nine the differ- ence is 0.20 per cent or more; in three the difference is over 0.30 per cent, and in one the difference amounts to 1.18 per cent. Lime can be deter- mined so that duplicates run within 0.05 per cent of each other; therefore it appears that in twenty-seven out of thirty-two cases the method of calculating lime from the results for the various forms of sulphur affords no check whatever on the true amount present. In Table III of Michigan Technical Bulletin No. 6, out of fifteen samples, nine have differences of 0.06 or more between the calculated and deter- mined values for lime, while five have differences of 0.10 or more. It may be urged that the experimental errors in the determination of the different forms of sulphur may account for these discrepancies. Let us inquire more closely into this contention. Sulphate and sulphite sulphur 90 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 Results from Michigan Technical Bulletin No. 6. LIME (CaO) SAMPLE REFERENCE Determined Calculated Difference per cent per cent per cent 4A 7.16 7.08 +0.08 Table V, p. 10. 6A 6.92 8.10 —ikatks) Table V, p. 10 1A 8.02 7.84 +0.18 Table VII, p. 11 1B 9.00 8.80 +0.20 Table VII, p. 11. 2A 8.30 8.16 +0.14 Table VII, p. 11. 2B 8.44 8.43 +0.01 Table VII, p. 11. 3D 8.85 8.58 +0 .27 Table VII, p. 11. 1A 7.18 7.07 +0.11 Table VIII, p. 12. 1B okt 7.06 +0.05 Table VIII, p. 12. 1C 7.09 7.30 —0.21 Table VIII, p. 12. 2A 6.97 6.96 +0.01 Table VIII, p. 12. 2B 6.88 6.77 +0.11 Table VIII, p. 12. 2C 7.10 7.13 —0.03 Table VIII, p. 12. 3A 9.83 9.50 +0.33 Table VIII, p. 12. 3B 8.93 8.77 +0.16 Table VIII, p. 12. 3C 9.54 9.46 +0.08 Table VIII, p. 12. 4A 8.85 8.41 +0.44 Table VIII, p. 12. 4B 8.65 8.39 +0 .26 Table VIII, p. 12. 4C 8.67 8.72 —0.05 Table VIII, p. 12. 5A TA9 7.65 +0.14 Table VIII, p. 12. 5B 7.95 7.76 +0.19 Table VIII, p. 12. 6A 8.21 8.00 +0.21 Table VIII, p. 12. 6B 8.07 8.03 +0.04 Table VIII, p. 12. 7A 5.77 5.61 +0.16 Table VIII, p. 12. 7B 5.66 5.58 +0.08 | Table VIII; p. 12. 7C 5.79 5.84 —0.05 Table VIII, p. 12. 8A 6.52 6.24 +0 .28 Table VIII, p. 12. 8B 6.16 6.07 +0.09 Table VIII, p. 12. 8C 6.31 6.16 +0.15 Table VIII, p. 12. 3D 8.89 8.81 +0.08 Table IX, p. 14. 4D 8.51 8.50 +0.01 Table IX, p. 14. 8D 6.20 6.29 —0.09 Table IX, p. 14. are present only in traces and are generally determined within 0.01 per cent, which would affect the value of lime calculated therefrom 0.01 1.748=0.017 or 0.02. In many eases, as in sample No. 3, Table III (Mich. Tech. Bul. No. 6), there is no sulphate or sulphite sulphur present and yet in this case there is a difference of 0.13 between the calculated and determined values for lime. In general, then, the error in the de- termination of sulphur in the form of sulphite and sulphate can affect the calculated value for lime but little. In the determination of thiosulphate, every 0.10 ec. n/10 I=0.00064 gram thiosulphate sulphur=0.00056 gram CaO. Therefore, for every error of 0.10 ce. of tenth-normal iodin made in the titration for thiosulphate, the percentage of lime calculated there- from would be affected by 0.14 per cent, assuming that 0.4 gram concen- trate containing 10 per cent of lime is under titration. But in the iodin titration method, any error in the monosulphur equiva- lent titration affects the thiosulphate titration. Thus, if 0.10 ec. of tenth- 1915] ROARK: ANALYSIS OF LIME-SULPHUR SOLUTIONS 91 normal iodin too much is used for the first titration, the second titration will be short just that much. Now, 0.10 ce. n/10 I=0.00016 gram sulphid sulphur=0.00028 gram CaO. Thus, from the monosulphur equivalent titration, assuming 0.10 ec. of tenth-normal iodin too much to be used, 0.00028 gram of lime too much would be calculated, but the succeeding thiosulphate titration being 0.10 ee. short and 0.00056 gram of lime being calculated therefrom, there would be the result of running over 0.10 cc. of tenth-normal iodin in the monosulphur equivalent titration, 0.09028 gram of lime less, or 0.07 per cent less calculated lime on a 0.4 gram aliquot running 10 per cent lime. Assuming the errors to run the other way, too little monosulphur equiva- lent and too much thiosulphate, there would be for every 0.10 ce. of tenth- normal iodin that the monosulphur equivalent titration was short, 0.00028 gram of lime, or 0.07 per cent more calculated lime. An error of more than 0.10 ce. of tenth-normal iodin in these titrations would be excessive provided the precautions emphasized by the associate referee, namely, small dilution, and the use of nickel sulphate and starch paste as indicators, are observed, therefore, experimental errors in the determination of the various sulphur compounds (assuming only the poly- sulphid, the thiosulphate, the sulphite and sulphate to be present) are insufficient to explain the large errors in the values for calculated lime figured therefrom. The logical conclusion, therefore, is that other sulphur compounds of calcium are present. How the presence of these other bodies will affect the values for calculated lime will now be shown. WHY THE VALUES FOR DETERMINED LIME ARE LOWER THAN THE VALUES CALCULATED FROM RESULTS BY THE IODIN TITRATION METHOD, BUT HIGHER THAN THE VALUES CALCULATED FROM RESULTS BY THE ZINC CHLORID METHOD. Lime (CaO) in commercial concentrates calculated from results by the iodin titration method are higher, while lime calculated from results by the zine chlorid method are lower than the determined values for lime. The explanation for this is as follows: The complex oxysulphid compounds which have been found by several observers to be present in pure lime-sulphur solutions are generally repre- sented by the following formulae: 2CaO-CaS;-H:O; 3CaO-CaS,12H.0O; 4Ca0O-CaS,18H20; 3CaO-CaS8;3-H.0. In titrating a lime-sulphur solution, the iodin would react with one of these compounds as follows: 4CaO-CaS,18H,0-+ I, = Cal.+48+4Ca(OH)2+ 14H20 92 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 This reaction would take place while the monosulphur equivalent titra- tion was being made. The calcium hydroxid would then react with iodin according to the equation: 2Ca(OH)2+4I =Ca(OI)2+Cale+2H2O This reaction between calcium hydroxid and iodin would probably take place both in the monosulphur equivalent titration and in the thiosulphate titration.! According to this equation 2 [=1CaO, but according to the reaction between calcium thiosulphate and iodin, 2 [=4 thiosulphate S; and since thiosulphate 8 times 0.874= CaO combined with it as calcium thiosulphate, 2 1=4X0.874=3.5 CaO. Therefore, for every molecule of caleium hy- droxid titrated in the thiosulphate titration with iodin, 3.5 molecules will be calculated therefrom and reported according to the iodin titration method of analysis. If the free calcium hydroxid is titrated as mono- sulphur equivalent, for every molecule present 1.75 molecules will be eal- culated. Assuming that as much free calcium hydroxid is titrated as monosulphur equivalent as is titrated as thiosulphate sulphur, the fol- lowing is true: For every molecule of lime but one of the complex oxysul- phid compounds present in the lime sulphur solution, 2.63 molecules of lime will be calculated when the sample is analyzed according to the iodin titration method. It is interesting to note the result on this complex oxysulphid com- pound when the solution is analyzed by the zine chlorid method. On addition of the ammoniacal zine chlorid, all the sulphur of the compound, which is present as sulphid sulphur, will be precipitated as zine sulphid or polysulphid. It may be that the whole compound will be precipitated, but at any rate all the sulphur will be. Assuming that the free calcium hydroxid does pass into the filtrate from the zine sulphid precipitate be- fore titration with iodin for thiosulphate sulphur all the calcium hydroxid, together with the ammonia, will have been changed to chlorid in the neu- tralization with hydrochloric acid, and will not, therefore, interfere with the titration for thiosulphate sulphur according to the zine chlorid method. In the zine chlorid method, this extra calcium hydroxid would not be titrated and so would not be calculated at all, but would appear in the gravimetric determination, thus making the determined value higher than the calculated. The presence of free calcium hydroxid would tend to affect results in the same manner as the complexes just mentioned. The presence of hydrogen sulphid would tend to make the results for mono- sulphur equivalent high and the value of lime calculated therefrom cor- respondingly high. The presence of hydrosulphid or hydroxyhydro- 1 Bur. Chem. Bul. 162, p. 41. 19165] BOARK: ANALYSIS OF LIME-SULPHUR SOLUTIONS 93 sulphid would tend to raise the results for thiosulphate sulphur as shown by McDonnell! and, of course, the value of lime calculated therefrom. From a study of the tables of results of analysis of commercial concen- trated lime-sulphur solutions on page 87 it is seen that the theory of this explanation is verified in every instance by the facts. CONCLUSIONS. (1) It is shown that lime-sulphur solution is a complex solution, con- taining many compounds, the presence of which is generally overlooked. (2) It is shown that in the zine chlorid method all compounds that can interfere with the titration of thiosulphate sulphur (excepting, of course, traces of sulphite) are precipitated by the addition of ammoniacal zine chlorid solution, or are rendered innocuous by the neutralization of the filtrate with hydrochloric acid. (3) It is shown that in the iodin titration method such compounds will markedly affect the results obtained in the titrations, more especially in the titration for thiosulphate sulphur. (4) A theory is advanced explaining the discrepancies observed in the determined and in the calculated values for lime, whether calculated from results obtained by the iodin titration or by the zine chlorid method. RECOMMENDATIONS BY THE ASSOCIATE REFEREE. I. Lime-Sulphur Solutions. It is reeommended— (1) That the method for total sulphur be changed to read as follows: Weigh accurately 10 grams of the solution and make to 250 ec. with carbon-dioxid- free water. Transfer a 10 cc. aliquot to a 400 cc. beaker, add about 3 grams of sod- ium peroxid, cover immediately with a watch glass and warm on the steam bath, with frequent shaking, until all the sulphur is oxidized to sulphate, adding more sodium peroxid if necessary. (Instead of sodium peroxid, hydrogen peroxid may be employed, in which ease first add 3 cc. of a concentrated (1:1) solution of sodium hydroxid. In either case carefully test all reagents for sulphur, and if present, make corrections accordingly.) Dilute, acidify with hydrochloric acid, evaporate to dryness and filter to remove silica. Dilute filtrate to 300 cc., add 50 ec. of con- centrated hydrochloric acid (see J. Amer. Chem. Soc., 1911, 83: 844), heat to boiling, and precipitate, stirring constantly with a 10 per cent solution of barium chlorid. This should be added at sucha rate that about 4 minutes are required in running in the amount necessary (11 ce. for 1 gram of barium sulphate); the rate is best regulated by attaching a suitable capillary tip to the burette containing the barium chlorid solution. Evaporate the whole to dryness on the steam bath (this may be done immediately after precipitation), take up with hot water, filter through paper, 1 Bur. Chem. Bul. 162, p. 41. 94 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 wash until the washings are free from chlorid, ignite very carefully so as to obviate reduction, and heat to constant weight over a Bunsen burner. Calculate sulphur from the weight of barium sulphate, using the factor: Wt. BaSO, X 0.13738 = Wt. S. (2) That the method for sulphid sulphur be changed to read as follows: Dilute 25 ec. of the solution, prepared as for total sulphur, to about 100 cc. and add ammoniacal zine chlorid solution (prepared by dissolving 50 grams of pure zine chlorid in water and adding ammonia in sufficient quantity to redissolve the precipitation first formed) until the sulphid is all precipitated, as will be shown by adding a drop of the clear solution to a few drops of nickel sulphate solution. Filter immediately, wash precipitate thoroughly and transfer, together with the filter paper, to a beaker. Cover with water, disintegrate with a glass rod and add about 3 grams of sodium peroxid, keeping the beaker well covered with a watch glass. Warm on the steam bath with frequent shaking until all the sulphur is oxidized to sulphate. (In case hydrogen peroxid is used as the oxidizing agent treat filter containing zine sulphid with a concentrated solution of sodium hydroxid, warming on the steam bath, before adding the peroxid.) Make slightly acid with hydro- chloric acid, filter to remove filter paper, wash thoroughly with hot water, and determine sulphur in the filtrate exactly as given under total sulphur. (8) That the method for thiosulphate sulphur be changed to read as follows: Dilute 50 ce. of the solution prepared as for total sulphur to about 75 to 100 cc. in a 200 cc. graduated flask. Add ammoniacal zine chlorid until in slight excess and make to mark. Shake thoroughly and filter through a dry filter. To 100 cc. of the filtrate add Methyl Orange and exactly neutralize with dilute hydrochloric acid. ‘Titrate this solution with standard iodin solution (approximately twentieth- normal), using a few drops of starch paste as indicator. From the number of cubic centimeters of iodin solution used, calculate the thiosulphate sulphur present. The value of the iodin solution being given in terms of arsenic trioxid, use the fac- tor: As.O3; X 1.29628 = thiosulphate 8. (4) That the method for sulphate sulphur be changed to read as follows: To the solution from the determination of thiosulphate add 2 or 3 drops of hydrochloric acid, precipitate in the cold with barium chlorid solution, allow to stand over night, filter, and from the weight of barium sulphate calculate sulphur and report as sulphate sulphur. (5) That total lime be determined as follows: To 25 ce. of the solution prepared as for total sulphur, add 10 cc. of concentrated hydrochloric acid, evaporate to dryness on the steam bath, take up with water and a little hydrochloric acid, warming until all the calcium chlorid is dissolved and filter from sulphur and any silica that may be present. Ignite filter containing fil- tered sulphur, take up in a little dilute hydrochloric acid, and filter through a small filter into original filtrate. Oxidize filtrate by boiling with a little concentrated nitrie acid, make ammoniacal, filter from iron and aluminum if present, heat to boiling and precipitate with ammonium oxalate solution. Filter and heat to con- stant weight by means of the blast lamp. : 19165] AVERITT: ANALYSIS OF A LIME-SULPHUR SOLUTION 95 A SHORT METHOD FOR THE ANALYSIS OF A LIME-SULPHUR SOLUTION. By S. D. Averirt. The writer has found that an analysis of a lime-sulphur solution suffi- ciently accurate for commercial work may be made in about two hours. This is accomplished by weighing directly the precipitated sulphur from the iodin titrations as given in the report on insecticides this year, and estimating the sulphate and sulphite sulphur as precipitated in the iodin methods by the degree of turbidity. If the sulphate and sulphite sulphur are precipitated in not more than 50 cc., the estimation after a little prac- tice is quite accurate, never differing more than 0.02 or 0.03 per cent from the determined value. The most rapid and accurate method of weighing the sulphid sulphur is as follows: Wash a7 em. ashless filter several times with suction, dry thoroughly in the water oven; put in a weighing bottle and weigh at once. Filter the sulphid sulphur on this weighed filter, washing thoroughly, drying and weighing in the bottle as before. The difference in weight is the sulphid sulphur. During the first drying of the filter the titrations with iodin are made, 2 or 3 drops of dilute hydrochloric acid added, the solution placed on the waterbath for a few minutes (do not heat above 40° or 50°C). When sulphur collects, it is ready to be filtered. To the filtrate and one washing (the whole should not be over 40 or 50 cc.) add 2.5 ec. of 10 per cent solution of barium chlorid, shaking thoroughly. Compare the turbidity with a standard solution in the same volume. The following table shows the comparison of the sulphid sulphur weighed directly and determined as barium sulphate on the present association samples: Comparison of sulphid sulphur weighed directly and as barium sulphate. SAMPLE 1 WEIGHED AS SAMPLE 2 WEIGHED AS Sulphur Barium sulphate Sulphur Barium sulphate 5.83 5.82 5.86 5.88 5.83 5.89 5.92 5.92 5.78 5.77 5.80 5.87 ee a 5.80 ae Average, 5.81 5.82 5.84 5.89 On a concentrate the sulphid sulphur by difference was 23.22 per cent; weighed as sulphur 23.22 per cent and 23.14 per cent, average 23.18 per cent. 96 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 The writer has found this method of weighing the sulphid sulphur just as accurate as weighing it as barium sulphate and the total sulphur as the sum of the sulphid, thiosulphate, and sulphate sulphur never differs over 0.10 or 0.15 per cent from the total sulphur as determined, and the time saved outweighs the small loss of accuracy. SECOND DAY. TUESDAY—MORNING SESSION. REPORT ON WATER. By W. W. Skinner, Referee. The work on water analysis for the past year was confined to a study of the proposed methods for strontium, iodin, and bromin. A sample was sent to each of nine chemists and reports have been received from six, three of whom represent the Water Laboratory of the Bureau of Chemis- try. The sample was known to contain in the aliquot taken for analysis approximately 160 mg. of calcium, 31 mg. of strontium, 0.25 mg. of iodin, and 3.25 mg. of bromin. According to the method, the differences between the total oxids and the strontium oxid calculated from the stron- tium sulphate determined, is the value for calcium oxid. The results reported are very satisfactory. Results on water analysis. (mg.) CALCIUM BY ANALYST ae PACA hs aa Rey ee *a) eRe 1 160.00 157.93 119 90 10.07 2.68 159.40 156.51 118.90 On | areee te 2 Ho) |) SE5e8 28.10 0.21 2.00 3 158.10 149 .36 31.10 0.15 3.59 157.70 147 .65 31.80 0.15 3.20 Pee iin ek oy, <5) ae ee 0.14 3.36 4 161.90 160.50 28.10 0.21 2.21 160.50 160.00 30.10 0.23 2.35 159.20 15S 000R" |e SM Set ne a 5 159.70 157.40 30.70 0.14 19 30 159.40 157.70 31.20 0.15 19 30 158.20 157.80 32.60 0.15 10.35 6 163.00 159.40 199 40 0.22 1.95 160.30 158 .90 122.50 0.22 2.48 conc eee ane | eee 0.25 i Average 159.78 156.76 30.46 0.18 2.65 Theory...... GOROOWN meee: 31.00 0.25 3.25 1 Omitted from average. The values, as you will note in the accompanying table, under the head- ing “Calcium by difference,’”’ agree very well with theory. This seems to hold true irrespective of whether the strontium is low or not. Two 97 98 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMIsTs [Vol. J, No. 1 analysts, Nos. 1 and 6, report low results on strontium, but correct results for calcium, indicating that the probable error in the method is due to excessive washing of the oxalates causing a loss due to the solubility of calcium and strontium oxalate, especially if hot water is used; and as the strontium oxalate is the more soluble of the two, the loss is first apparent in the low results for strontium. Hildebrand! in his methods of rock analysis reports some work by Richards and others bearing on this and remarks that this solubility of the oxalates “needs greater attention than it ordinarily receives from analysts.’”’ If double precipitation of the oxalates is made as it should be, it is surprising to most analysts to learn how little washing is really necessary. The column headed “Calcium direct” gives the determination of calcium after the separation of strontium. This figure has been regarded heretofore as a check on the calcium by difference, but it should be noted that these results are invariably low. The exact cause has not been determined, but it is probably due to the ether-alcohol mixture failing to dissolve all of the calcium nitrate. An interesting observation, however, is in such cases where the calcium is low by direct determination, the strontium is not necessarily high. This is no doubt due to the fact that strontium sulphate is precipitated out of the alcoholic solution, while the contaminating calcium which the ether-alcohol failed to extract remains in the solution and does not come down as sulphate. With the low reports by two of the analysts on strontium thus accounted for, the re- sults for calcium and strontium are considered to be quite satisfactory. I have thought it advisable, however, not to recommend the adoption of the method, but to refer the matter to the referee for next year with the suggestion that further work be done with the idea of determining especially the minimum amount of washing necessary for the oxalate precipitate. Three of the six analysts obtained results on iodin that were quite satisfactory. One result was very much and two others slightly lower than the amount known to be present. The results on bromin are only fairly satisfactory, but indicate that with care in the manipulation and experience in the determination, satisfactory results may be obtained. The colorimetric methods for iodin and bromin have given results in the laboratory of the referee that are considered satisfactory for quantities of iodin and bromin such as are usually found in waters. It is true, how- ever, that considerable experience in the manipulation is necessary in order to get satisfactory results, especially in the case of the determination of bromin where the use of an excessive amount of chlorin water will give discordant results. Unsatisfactory results are also obtained if too much 1U. 8S. Geol. Sur. Bul. 422. 1915] SKINNER: WATER 99 of the original solution is taken so that the carbon bisulphid contains sufficient bromin to give a distinct red color. Fading of the color under such conditions is very rapid, especially if an attempt is made to make the readings in comparison tubes. Under such conditions, and in fact gener- ally, it has been found that for bromin the comparison can best be made in small Erlenmeyer flasks without removing the solution from which the bromin is extracted. By using a definite amount of carbon bisulphid and shaking the flask in such a manner as to bring it together in one large globule the comparison in the flask may be made quite satisfactorily. It is evident also from the results reported and from work done in our laboratory that one serious difficulty in applying the method is due to loss in extraction of the residue with alcohol, especially a loss of bromin. When the residue is not too large it is possible that satisfactory results may be obtained on the residue direct without extraction. In the Bureau of Chemistry laboratory quite satisfactory qualitative results are ob- tained on the water directly. While the referee is satisfied that satis- factory results may be obtained by the colorimetric method for iodin and bromin, he is of the opinion that it would be unwise as in the case of strontium to recommend these methods for adoption as official until fur- ther work has been done. It is, therefore, recommended that the referee for the next year be directed to continue the work on methods for the determination of strontium, iodin and bromin along the lines suggested. The methods proposed two years ago and which were given their first reading for final adoption at the last meeting come up automatically for final adoption at this time. It is, therefore, recommended that the proposed methods, except those for strontium, iodin, and bromin be adopted as official. I desire to report that the referee has taken part in the discussions of a joint committee representing the American Chemical Society, the American Public Health Association, and this association for the purpose of bringing about a greater uniformity of methods for water analysis. It has been tentatively agreed upon that the methods for mineral analysis would be considered by this association, that methods for technical analy- sis would be considered by the committee of the American Chemical Soci- ety, while the sanitary and bacteriological analysis would receive special attention from the American Public Health Association. This general assignment of work in addition to unifying methods, is for the purpose of avoiding duplication of effort. The arrangement is only tentative, of course, and does not prevent any one of the organizations from doing work in any of the particular phases of water analysis mentioned. 100 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. J, No. 1 REPORT OF COMMITTEE A ON RECOMMENDATIONS OF REFEREES. By B. B. Ross, Chairman. (Phosphoric acid, nitrogen, potash, soils, inorganic plant constituents, insecticides, water.) PHOSPHORIC ACID. It is recommended— (1) That further work be done on methods for determining total - phosphoric acid in basic slag. Approved. (2) That further work be done on methods for determining available phosphoric acid in basic slag. Approved. (8) That further attention be given to the presence of iron and silica in the magnesia precipitate and that methods designed to eliminate these substances be studied. Approved. (4) That the methods outlined for total and available phosphoric acid be tried out with a synthetic solution representing as closely as possible a solution of the average basic slag. Approved. (5) That the referee for next year study the titration method for making neutral ammonium citrate solution (J. Ind. Eng. Chem., 1918, 5: 567). Approved. - NITROGEN. It is reeommended— (1) That the ferrous-sulphate-zinc-soda method be adopted as a provisional method and that it be further studied during the coming year. Approved for final action as provisional in 1914. (2) That the alkaline permanganate and neutral permanganate methods for organic nitrogen availability (Bul. 162, p. 13) be adopted as official methods and that they be printed under the respective designations of “Method for the determination of organic nitrogen soluble in alkaline permanganate” and ‘‘ Method for the determination of organic nitrogen soluble in neutral permanganate.” Approval for final action as official in 1914. (3) That the referee for next year study the Kjeldahl-Gunning-Arnold method and modifications thereof with a view to its employment for the determination of nitrogen in fertilizers and fertilizer materials. Approved. 1915] ROSS: REPORT OF COMMITTEE A 101 POTASH. It is reeommended— (1) That further codperation be secured in testing (1) the use of denatured alcohol for washing K.PtCl,, with special reference to the denaturing agents, (2) the necessity for the use of hydrochloric acid in the water extract in potash determinations. Approved. (2) That further work be done on the perchlorate method. Approved. (3) That the study of potash availability be continued another year and that the effect of decomposing green material also be studied. Approved. SOILS. It is reeommended— (1) That the work on humus be discontinued and that the official method for humus and humus nitrogen be eliminated from the revised methods of analysis. Approved. (2) That methods for determining organic carbon and nitrogen in soils be studied during the coming year. Approved. (3) That further study be made of methods for obtaining aqueous soil extracts and that the reduction method for nitrates and the methods for nitrites and ammonia, as adopted for waters, be approved for soil solutions. Approved. INORGANIC PLANT CONSTITUENTS. It is reeommended— (1) That no further work be done on the Schreiber method for total sulphur. Approved. (2) That the official method for iron and aluminum be extended to include calcium and magnesium in the presence of minute quantities of manganese. Approved. INSECTICIDES. It is reeommended— (1) That the whole question of the analysis of lime-sulphur solutions be referred to a committee of review, consisting of W. F. Hillebrand, C. 8. Catheart, and H. H. Hanson, whose duty shall be to consider the reports of the work of the past four years and to undertake such additional in- vestigations as may be necessary to reach a conclusion as to the merits 102 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 of the proposed methods, a report of their findings to be rendered at the next meeting. Approved. (2) That the method of digestion for water-soluble arsenic in lead arsenate (p. 75 of referee’s report) be made a provisional method. Approved. (3) That the method published by C. C. Hedges of Cornell Uni- versity (J. Ind. Eng. Chem., 1909, 1: 208), for the determination of arseni- ous oxid in Paris green and other insecticides, be compared by the next referee with the official method now in use. Approved. ° (4) That the Lloyd method for nicotin in tobacco and tobacco ex- tracts (p. 76 of referee’s report) be compared with the official method (Kisslings). Approved. WATER. It is reeommended— (1) That the referee for next year be directed to continue the work on methods for the determination of strontium, iodin, and bromin along the lines suggested in the report of the referee. Approved. (2) That the methods proposed in Circular 108, except those for strontium, iodin, and bromin, be adopted as official. Adopted, final action. REPORT OF COMMITTEE ON AVAILABILITY OF PHOSPHORIC ACID IN BASIC SLAG. By C. B. Witirams, Chairman. Since the last meeting of the association the Committee on Availability of Phosphoric Acid in Basic Slag has held two meetings, one in Atlanta on November 14 and 15, and one in Washington on January 24, and has spent much time in outlining a satisfactory plan for conducting pot experiments to study the availability of the phosphoric acid of basie slag as compared with that contained in sodium acid phosphate, double super- phosphate, acid phosphate, and finely-ground phosphate rock. During April, the directions and phosphatic materials for the pot work were forwarded to fourteen station workers who had previously indi- cated that they would be in a position to codperate with the committee in the investigations. Most of these have started the work, but only a 1 Presented by H. D. Haskins. 1915] WILLIAMS: AVAILABILITY OF PHOSPHORIC ACID 103 few, as might be expected, have been able to secure results this early. At present seven station workers are co6perating in the field experiments, an outline of which was submitted at the last meeting of the association. Other workers have indicated that they would be in a position later to take up co6perative work with the committee. It is suggested that if possible the referee on phosphoric acid use for his basic slag studies the same slags that are being used by this committee. By thus doing it is felt that the results secured by each may prove mutu- ally valuable. SOURCE OF MATERIALS. At the beginning of the investigations the committee, after much correspondence, has in its judgment secured enough of each of the different phosphatic fertilizing materials to be used in the availability studies to provide all those coGperating in the pot and in the field experiments with the same materials throughout the entire period to be covered by the investigations. It was the idea of the committee to secure for these studies lots representative of the slags that are finding their way into our markets from different European and Canadian sources. The slags and other phosphatic materials finally selected and secured for the work were of the make and source indicated: Slag A. Manufactured by the Chemical Works of the Late H. and A. Albert, England, and secured through the Coe-Mortimer Company, New York, N. Y. Slag B. Manufactured by the Anglo-Continental Guano Works, Antwerp, Belgium, and secured through the Nitrate Agencies Company, New York, N. Y. Slag C. Manufactured by the Dominion Iron and Steel Works, Canada, and secured through the Cross Fertilizer Company, Sidney, Nova Scotia. Slag D. Manufactured in Belgium by a different firm from that manufacturing Slag B; secured through H. J. Baker and Brother, New York, N. Y. Ground blue phosphate rock. Mined at Gordonsville, Tenn., and secured through the Robin Jones Phosphate Company, Nashville, Tenn. Acid phosphate. Manufactured by and secured through the Caraleigh Phosphate and Fertilizer Works, Raleigh, N. C. Double superphosphate. Manufactured by the American Agricultural Chemica Company, Boston, Mass., and secured through the Coe-Mortimer Company, New York, N. Y. FINENESS AND COMPOSITION OF MATERIALS. These materials secured for the coéperative experiments had the fol- lowing average fineness and composition: 104 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMIsTS [Vol. I, No. 1 Fineness and composition of materials. DENIES) PERCENT- | PERCENT- ————eeeeeeeSSssSsssSs| AGE OF AGE OF San La- | PERCENT- rr rie gee ee ees ere Lee or PHORIC PHORIC 1 mm. 4 mm. 4 mm. yy IMM. ate AcrD? Slag As oo cassccienee 99.86 | 99.31 | 97.28] 65.36} 18.06) 15.87] 0.19 SIlapiB yaa ter rere 99.81 | 98.42 | 95.01 | 73.17] 17.84] 14.74] 0.29 Slagi@ ess. eeeces ees 99.52 | 98.14 | 94.14] 68.43] 13.03] 18.25] 0.28 Slag eee oe 99.16 | 96.50} 91.10] 67.34] 15.57] 14.98] 0.30 Phosphate rock......} 99.65 | 99.45] 95.80] 72.82} 29.40} ..... Neidiphosphatere.- cel) y-raee le meecerel | eco tee ol|ateieiaeee 19.49 | 17.82 Sodiumpphosphatesc|) sii) setceealll peicies> aabon || Zeer || ZOeeye Double superphos- DRALC os teseea eco al Pe elace ill mice eeee ell: secee 46.25 | 46.02 1 Available phosphoric acid determined in slags by Wagner’s 2-per cent citric acid method and in other materials indicated by the official ammonium citrate method. DIRECTIONS FOR CONDUCTING POT EXPERIMENTS. NATURE OF SOIL. If the field work is being undertaken at the same time, use for the pot experi- ments the surface soil after the phosphoric acid has been exhausted in the field by growing of crops. Fill all pots with well-mixed soil taken from a number of places on Plats 5 and 34 (see fig. 1 opposite, p. 50 of Bur. Chem. Bul. 162); those who have access to fields that are known to be deficient in phosphoric acid may use such soils in place of what is recommended. The experimenter should prove beyond a doubt that the soil is deficient in phosphoric acid, even though this had been indicated in previous field tests. The thorough breaking up and mixing of the soil in prepar- ing it for pot work, together with the liberal quantity of lime which is later recom- mended, will unquestionably render available some of the inert phosphoric acid in the soil. In such a soil it probably would not be as difficult to deplete the phos- phoric acid as would be the case if the soil were taken from a field where no previous efforts had been made to exhaust the soil of this constituent. Some time might therefore be saved in choosing a soil which was deficient in phosphoric acid. The experiment is to be conducted with the same soil and is to be carried on in two main divisions. In the first, some leguminous crop such as clover, vetch, or cowpeas, is to be grown without fertilization (except lime), the fertilizer being applied subsequently in intimate contact with the green crop after it has been passed through a cutter and mixed with the soil in the respective pots. In the second division the soil in the pots is to remain in a fallow condition, after an appli- cation of lime, during the growth of the legume. In one of the divisions the legume is grown and turned under in order to study what is thought to approach more nearly practical soil requirements for best farm results with some of the phosphate materials under investigation. Subsequent to the production of the legume selected for growth in the first division of pots during the preliminary period, the soil of all the pots of both divisions of the experiment should be treated at the same time and in the same manner. The seeding of the legume should be uniform for all the pots that are to grow this crop. 1915] WILLIAMS: AVAILABILITY OF PHOSPHORIC ACID 105 TAKING SOIL. Take soil for the pot experiments to the depth of 7 inches, excluding stubble and other undecomposed organic material and stones. On a sample of the mixed soil as weighed into the pots, determine the loss upon air-drying, pass the air-dried soil through a 3 mm. sieve, determine the percentage of coarse material, select a 2-quart sample from the mixed, sieved material and dry it for 8 hours at 70°C., determining the loss of water. Preserve for possible future analysis. FILLING POTS. Fill the pots from the thoroughly-mixed soil, placing equal quantities by weight in each pot so that after the soil has been compacted its surface will be within an inch of the top. Those who use pots over a foot in depth may, if they prefer, fill the excess depth at the bottom with subsoil, or they may, in case of pots of any depth, use the surface soil on top to the depth at which it occurred in the field and use subsoil for the remainder at the bottom. The lime and fertilizer applications should be mixed with the surface soil and should be made on the basis of air-dry surface soil, exclusive of subsoil. If possible, all pot tests in the four series should be run in duplicate. APPLICATION OF LIME. The lime should preferably be added on the basis of analytical data secured in accordance with the Veitch method (J. Amer. Chem. Soc., 1902, 24: 1120-1128; Bur. Chem. Bul. 73, p. 136) using for this purpose a portion of the oven-dried sample already mentioned, having first passed this portion through a 1 mm. sieve. De- termine the relation of this sieved subsample to the surface soil as weighed mto the pots, as well as its relation to the main oven-dried sample. Add per air-dried soil, in excess of the lime requirements as determined by the Veitch method, calcium carbonate at the rate of about 0.10 per cent for L and 0.15 per cent for Lx. APPLICATION OF NITROGEN. For applications of nitrogen add for air-dry soil, 0.06 per cent of nitrogen from dried blood and 0.01 per cent from nitrate of soda, for N, and 0.09 per cent of nitro- gen from blood and 0.015 per cent nitrogen from nitrate of soda, for Ny. APPLICATION OF POTASH. For the application of potash, add for air-dry soil, 0.10 per cent of potash for K, and 0.15 per cent of potash for Ky, of low grade sulphate of potash. APPLICATION OF PHOSPHORIC ACID. For the application of phosphoric acid, add for air-dry soil, 0.007 per cent of phosphoric acid for P;, 0.014 per cent of phosphoric acid for P, 0.021 per cent of phosphoric acid for Pj,, and 0.028 per cent of phosphoric acid for Pz of the different phosphatic materials indicated. FERTILIZERS. If, during the growth of the plants, it appears that those which receive sodium phosphate or double superphosphate together with Ny; and Kj are growing better than those with N and K, all pots which received N and K should have added to them an equal amount of a solution of nitrate of soda and low grade sulphate of potash and those which had received Ny; and Ky, 50 per cent more than this, since it is necessary in an experiment of this kind to have the N and K present in optimum amounts. 106 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 Owing to the varying requirements of different soils it is, of course, only possible for the committee to suggest what should be added. Those who know the specific requirements of their soils, should feel free to vary the applications suggested. It should be borne in mind, however, that in pot experimentation, especially if the plants are allowed less feeding area than in the field, the applications must be very liberal. It is realized that the requirements of the different kinds of crops are not the same, but it is hoped that the applications suggested above are liberal enough to cover each case. It is especially difficult for the committee to suggest applications of phosphoric acid suitable under different conditions, since it is desired to have the applications of P; in sodium phosphate and double superphosphate below the optimum requirements of the plants, as indicated by the larger crop which should be secured by P and Py. The suggested applications of lime are liberal in order to remove the possibility of the basic material of the slags exerting any beneficial influence either on the re- action of the soil or in connection with the liberation of plant food ingredients. It is advisable to mix fertilizing materials with the soil, then moisten thoroughly and uniformly and allow to stand for two weeks in advance of seeding. CROPS. For the field experiments Japanese millet and dwarf Essex rape have been selected. In the pot experiments millet or spring wheat and rape should be used. Add an equal weight of good seed to each pot, thin to the best plants, finally allowing not less than nine square inches to each rape plant and about four squares inches to each millet or wheat plant, making sure that in the case of each crop an equal number of plants are left in each pot. WATERING THE PLANTS. Owing to varying conditions it is impossible to make suggestions regarding the watering except to call attention to the fact that at a given time approximately equal amounts of water should be added to plants of similar size. HARVESTING. Report the following data: Crops used; character of soil; depth of surface soil in the field; per cent of calcium carbonate required, according to the Veitch method, by the dry soil finer than 1 mm.; per cent of this fine soil in the oven-dried soil finer than 3 mm., and in the soil used in the pots when air-dried; kind and dimension of pots; depth of soil and of subsoil in pots; weight of surface soil in air-dry condition per pot; per cent material coarser than 3 mm. in air-dry soil; time of planting; num- ber of plants per pot; time of harvesting; state of maturity; method of drying the crops; weight of crop from each pot; variations from the suggested directions. SUPPLY OF MATERIALS. The sodium phosphate, double superphosphate, acid phosphate, Thomas slag phosphate and ground (blue) phosphate rock together with the analyses of same, will be supplied by the committee to those codperating in the experiments. These analyses are to be the basis of the phosphatic materials. FERTILIZER APPLICATIONS PER POT. N, 0.06 per cent of nitrogen from dried blood and 0.01 per cent from nitrate of soda. 19165] WILLIAMS: AVAILABILITY OF PHOSPHORIC ACID 107 K, 0.10 per cent of potassium oxid from low grade sulphate of potash. P, 0.014 per cent of phosphoric acid respectively from the different phosphatic materials indicated. L, 0.10 per cent of calcium carbonate plus that required by the Veitch test. Lx, 0.15 per cent of calcium carbonate plus that required by the Veitch test. Run the pot experiments with each crop in two main divisions. After lime is applied and before fertilizer treatment begins with any crop, on the first division a leguminous cover crop is to be grown and to be turned into the soil. During this period, the second division is to remain fallow after adding lime. In the case of each division, the following fertilizer treatments are recommended for the respective groups, each of which is to run in four series: Group 1 NKL. NKL. NP,KL Phosphoric acid from Slag A. NP;KL Phosphoric acid from Slag B. NP;KL Phosphoric acid from Slag C. NP;KL Phosphoric acid from Slag D. NP;KL Phosphoric acid from acid phosphate. NP,;KL Phosphoric acid from ground (blue) phosphate rock. NP;KL Phosphoric acid from sodium phosphate. NPKL_ Phosphoric acid from Slag A. NPKL_ Phosphoric acid from Slag B. NPKL_ Phosphoric acid from Slag C. NPKL_ Phosphoric acid from Slag D. NPKL_ Phosphoric acid from acid phosphate. NPKL_ Phosphoric acid from ground (blue) phosphate rock. NPKL_ Phosphoric acid from sodium phosphate. NPKL, Phosphoric acid from sodium phosphate. NPKi;L Phosphoric acid from sodium phosphate. NPyKL Phosphoric acid from sodium phosphate. NP2KL Phosphoric acid from ground (blue) phosphate. NPKL Phosphoric acid from double superphosphate. NPKL Phosphoric acid from double superphosphate rock. ad NR OCOKOONAUAR WD [ol ol NORCO AN DUR w The complete pot experiments comprise the following four series. 1. Millet or wheat (spring) after previous growth of legume turned into the soil. 2. Rape after previous growth of legume turned into the soil. 3. Millet or wheat (spring) without previous ‘growth of legume, but after fallowing. 4. Rape without previous growth of legume, but after fallowing. If only one pot is allowed to a single treatment it will require not less than 88 pots for carrying out the entire scheme of experiments and it is very important to carry out the full line of investigations in duplicate. It is especially important in Groups 9, 16, 21, and 22 that at least two pots (and preferably three) be used for each treatment. A resolution was introduced by C. H. Jones on behalf of the executive committee that three associate referees be appointed as follows: (1) For the special study of the Kjeldahl method for the determination of nitrogen; (2) for the study of alkali in soils; and (3) on feed adulteration. The resolution was adopted. 108 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 REPORT OF COMMITTEE ON FOOD STANDARDS. By Wm. Frear, Chairman. The committee on food standards has been marking time this year, as it has been for a year or two past. In the whole history of the stand- ards work it has been made absolutely manifest that to secure a final adoption and general acceptance of any work for standardization all official interests that could in any way be concerned in the final use of these standards must have opportunity for conference in the prepar- ation of the standards. Such co-relation of the various officials interested and needing to use standards was for a time impracticable. While there was need for work it would be unfortunate to have it carefully done and yet but partly accepted or not accepted at all because of the lack of proper representation in the body charged with the work of making standards. So far as this association is concerned, this committee was fully author- ized, but as for some years we had been associated with interests that needed to be considered, and since it was impossible for us to secure for a time the active assistance from those other organizations, this committee has not attempted to go ahead with the standards work. There is now, I am pleased to say, some prospect of coéperative work such as ought to be done. Unfortunately I was not present at the conference at which that united action was made a subject of resolutions, and, therefore, I am unable to report in any official way from that body to this, but I assume that a presentation of that matter will be made. REPORT OF COMMITTEE ON EDITING METHODS OF ANALYSIS. By J. K. Haywoop, Chairman. I have never called a meeting of the committee on editing methods of analysis, and I am afraid it was an error on my part not to do so. It was my understanding that there was already a committee to take care of the revision of methods, and that a mistake had been made in appointing our committee. There was already a committee on recommendation of referees and revision of methods; I was the first chairman of that com- mittee four or five years ago, and at that time it was the idea of a good many that that committee was going to handle the revision. I think the committee on recommendation of referees and revision of methods is the proper one to revise the methods, because of its familiarity with the subject. Therefore, it was my thought that the appointment of this special committee was possibly a mistake on the part of the chair, so I did not take any action, and thought the subject would be referred to 19165] VAN SLYKE: STUDY OF VEGETABLE PROTEINS 109 the other committee. Now, the chair assures me that there is nothing in the resolution appointing the committee on recommendation of referees and revision of methods assigning to it the revision of methods, and if that is so, the committee on editing methods of analysis has made a mis- take in not handling the matter. If the association wishes this committee to go ahead with the revision, it will be glad to do so. REPORT OF COMMITTEE ON PRACTICABILITY OF ORGANIZING FOR STUDY OF VEGETABLE PROTEINS. By L. L. Van Stryke, Chairman. It will be useful to introduce this report with a brief historical state- ment. In 1901 the subject of separation of nitrogenous bodies was first added to the work of this association, and the first referee’s report made in 1902, at which time the subject was subdivided into three parts: Pro- teins of (1) milk and cheese, (2) plants, and (3) meat. In the ten years in which the subject of vegetable proteins has been a part of our regular work, no reports were made and no work done in the years 1908, 1909, and 1911; in 1903 and 1907 the reports consisted solely of summaries of literature; it was actually reported only five times. The separation of wheat proteins has formed the chief subject of study, with the exception of one year when attention was given to barley and malt. These state- ments obviously indicate that little advance has been made by the asso- _ciation in this field and the meagerness of results suggests that a complete change of policy must be adopted, if progress is to be made. Methods of study have undergone profound changes in recent years and workers in this field have a marked advantage over the workers of ten years ago. It is believed that the association should undertake anew a study of vegetable proteins, especially in relation to devising practicable methods of separation. While we do not think it necessary here to give detailed reasons for this belief or to attempt to point out more specifically the problems and methods of attack, the following suggestions are offered as a basis for organized effort. (1) In place of a single, transitory referee, there should be a small permanent committee, in which the personel shall remain as constant as circumstances permit. (2) Such committee should consist of members whose special train- ing and professional interests give them the highest degree of fitness and stimulation for the work. As chairman of this committee one name and only one suggests itself, that of T. B. Osborne, who is in a position of advantage to select other members. 110 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 (3) The permanent feature of the committee is absolutely essential to success in carrying out a systematic, effective line of work. (4) The committee should have full power in respect to planning and executing all details. (5) The suggestions of Mr. Osborne contained in the Report of 1912 (Bur. Chem. Bul. 162, pp. 154-159) furnish most valuable details for guidance in this work. REPORT ON FOOD ADULTERATION. By Juxtrus Hortvet, Referee. Recent years have witnessed many noteworthy changes both in respect to the purposes of chemical investigations affecting the control of foods and in the character of analytical methods of procedure. The enforce- ment of the Federal Food and Drugs Act, the extension of pure food legis- lation and regulation among the states, combined with the more intensive study of conditions of manufacture and distribution, have all contributed very largely, more than can be realized on momentary reflection, to a new era in that branch of applied analytical chemistry having directly under its charge the character of foods, drugs, and other articles of commerce. A reliable choice of an analytical method of procedure now shows a decided trend toward a greater refinement in details, a stronger emphasis, and a clearer insight respecting fundamental principles and true chemi- cal relations. It is now recognized more than heretofore that any really reliable chemical test or analytical procedure must be far beyond the elementary condition. Increased accuracy is demanded by the greater emphasis on exact definitions, true classifications, and the necessity of differentiating between products having close generic relations. There is evident a greater adaptation of methods of procedure to the nature of the material under investigation, a greater insistence on the recognition and handling of interfering conditions, and there is also a marked tendency withal toward increased reliability and a natural selection in the direction of shorter and simpler analytical procedures. The increased forensic demands of the past few years have influenced greatly the study of official methods of analysis and have stimulated the development of new methods in many directions. This has been illus- trated very strikingly with such products as vinegars, flavoring extracts, baking chemicals, canned goods, dairy products, brewed beverages, and coloring matters. There has been something akin to a revolution in the manufacturing, shipping, and marketing of food products, and the attend- ing changes have affected to a considerable degree the study and develop- 1915] HORTVET: FOOD ADULTERATION 111 ment of the analytical processes. Methods which only a few years ago served fairly well have now become inadequate or entirely obsolete. Such conditions as the increasing variety of vinegars made from various sugar wastes and sugar-containing substances, the identification of certi- fied and uncertified coal-tar dyes, the determination of the character of the mash from which a certain variety of beer is brewed, the compli- cations arising from the use of the homogenizer in the manufacture of milk and cream products, the changes attending cold storage of meats and eggs, and a multitude of other problems compared with which these illustrations are only a typical few have to be contended with now. The associate referees of the present year have shown to a degree comparing favorably with the work of preceding years, faithful attention to the study of proposed methods and modifications and to the further- ance of the wishes of the association as expressed at the last meeting. Details of the results accomplished will be adequately presented as the referees’ reports are read, but a few salient features of these reports may be appropriate. An examination of the report of the referee on heavy metals gives an impression of the progress now making toward accurate determinations in cases where metallic impurities are present in very minute quantities. It is impossible, however, to forego the suggestion that there is to some extent an overstepping tendency to experiment with colorimetric methods, notably in certain instances where obviously there can be little gained in respect to accuracy or rapidity of manipulation. These attempts, never- theless, are not to be discouraged. In connection with the report on lead, there is, so far as one can judge from the reports and comments of the collaborators, a prevailing tendency in favor of a gravimetric procedure. Too great importance cannot be attached to such work as that which has been carried out by the referee on water in foods. The systematic trying out of various dehydrating agents under different conditions is a line of work that is well worth attention and should be continued during the coming year. The same statement applies also to efforts on the part of collaborators on other subjects in which a similar investigation has been attempted. The referee on colors has undertaken a most extensive scheme of collab- orative work, a plan which apparently will lead to a complete revision of existing outlines of procedure. As a matter of valuable training, the detailed directions and outlines of the referee should be in the hands of all food chemists who are engaged especially on color work and who can arrange their time so as to carry out the plan of the work satisfactorily. In connection with the work on wine, it has been suggested that there is urgent need of more information regarding the composition and prop- 112 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 erties of the so-called “‘fixed acids,”’ with a view toward extending the abilities of the association beyond the handling of the ingredient which has heretofore been investigated as tartaric acid. The referee work on beer, though confined to a study of the methods for phosphoric acid, is very important. This report represents a neat piece of work and is a good model illustrating a kind of collaborative work much needed in order to perfect the methods. Accurate methods for beer analysis are also especially in demand now owing to the exigency which has recently arisen following the enforcement of regulations affect- ing the labeling of brewed or so-called malt beverages. A decided preference has developed in favor of the glycerin saponifi- cation method in carrying out the titer test in fats and oils. Also, a study of the Emery method for detecting beef fat in lard seems to have come to a favorable conclusion. If further work on this method is to be carried out, a study of comparative tests by the Belfield method is suggested both as a reliable check method and as a method which can be further perfected. The work on methods for meat and fish is noteworthy for the successful attempts which have been made to overcome unusual difficulties, par- ticularly in respect to the preparatory treatment of the material in the determinations of starch and sugar. The evolution of the Folin ammonia method is interesting and affords an illustration of what has been said regarding the tendency toward the adoption of shorter procedures and simpler details. The referee on flavoring extracts has brought out some interesting facts respecting the influence of certain interfering substances in the application of Folin’s test for vanillin. These considerations should be taken up and further study of the test is reeommended. The reports of analyses of authentic vanilla extracts constitute a valuable and very practical feature of the report, and this suggests the importance of con- tributing reliable data in all cases where such data may be available for the use of the committee on standards. The report of the referee on inorganic phosphorus covers two years of collaborative work, and for that reason affords a clearer view than would otherwise be possible of the considerable progress accomplished in this difficult field. This report is a good model of collaborative work, and while the subject-matter may not be directly classed under the general caption of Food Adulteration, the report is noteworthy, not only as a valuable contribution in its own field, but as one of a most helpful kind deserving a careful study on the part of all food analysts. The study of methods on inorganic phosphorus in plant and animal substances should be heartily endorsed and provision made for its continuation dur- ing the coming year. 1915] MATHEWSON: COLORS 113 The reports of the associate referees show the importance of experience as a qualification on the part of the collaborators. Too infrequently has it occurred that the same collaborators continue in a given line of work more than a year. Some arrangement is desirable whereby the referees may be assured of the continued assistance of experienced collab- orators, and to this end much improvement may result if trained analysts are encouraged to continue a line of study until they as well as the referees feel reasonably satisfied with the conclusion of an investigation. A large proportion of volunteering collaborators fail to report; this is in itself a matter of some concern, for, generally speaking, ‘‘The readiness of doing doth express no other but the doer’s willingness.”’ Quality of work more than quantity is however to be emphasized, and it is well to point out that many failures to report may be owing to an inadequate appreciation on the part of the volunteers of the true nature and importance of the collaborative work to be undertaken. In conclusion I wish to state my appreciation of the efforts on the part of the associate referees and their collaborators in the study of proposed methods and in the direction of new lines of investigation. There has been manifested a decided interest on the part of all and a hearty willing- ness to comply with the requests of the referee on food adulteration and the committee on recommendations of referees and revision of methods. We have in the majority of cases received copies giving full reports of work accomplished and in the main the abstracts which have been fur- nished have served their purpose well. In conclusion, thanking my asso- ciates for their interest and hearty coéperation, I am pleased to have these referee reports submitted to the association. REPORT ON COLORS. By W. E. Martuewson, Associate Referee. The work done on colors this season has comprised the examination by the collaborating analysts, of a number of samples of colored imitation liquors and further work by the associate referee on the solubilities and characteristic reactions of the common dyeing matters. The preliminary treatment of samples suspected of being colored con- sists usually in the extraction of the coloring matter with some solvent by which it is separated from most of the food material and obtained in a solution suitable for further examination. This gives no special diffi- culty with foods such as candies, sirups, and beverages, but with a few classes, as macaroni and oils, the methods for this step are still very un- satisfactory. This problem has not been taken up, at least directly, this year, as until the solubilities of the colors are better known, it can hardly be studied systematically. 114 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMIsTs [Vol. I, No. 1 COOPERATIVE WORK ON COLORED CORDIALS. Five samples of colored cordials were sent to the collaborating chemists, together with copies of the report of last year’s work and a set of tables designed to be supplementary to the procedure for examination of color mixtures and to provide the first draft for a systematic, coherent treat- ment of the subject. The essential features of this table are described in the latter part of this report. The samples were colored as follows, the percentage showing the amount of commercial dye in the mixture. No. 1. 0.01 per cent Ponceau 3R (S. & J. No. 56). 0.01 per cent Naphthol Yellow S (S. & J. No. 4). 0.01 per cent Orange I (S. & J. No. 85). (Contained about4 per cent of Orange II as subsidiary dye.) No. 2. Cochineal (S. & J. No. 706). 0.01 per cent Brilliant Scarlet (S. & J. No. 106). No. 3. 0.01 per cent Auramin (S. & J. No. 425). 0.01 per cent Rhodamia B (S. & J. No. 504). No. 4. 0.01 per cent Indigo Carmine (S. & J. No. 692). 0.01 per cent Palatine Scarlet (S. & J. No. 58). 0.01 per cent Eosin (S. & J. No. 512). No. 5. 0.01 per cent Guinea Green B (S. & J. No. 483). 0.01 per cent Metanil Yellow (S. & J. No. 95). About three weeks after mixture No. 4 was made, it was noted that the color had changed from bluish violet to orange red, and apparently in all cases this had taken place with the analyst’s samples before they had been examined. Most of the chemists used the table sent, together with personal notes and the various standard works. E. H. Grant reported that he had much trouble with the green dye in No. 5 and called attention to the fact that Auramin is decolorized and destroyed by caustic soda, a point not brought out in the table. The Guinea Green in mixture No. 5 was the commercial dye, which being produced by chemical reactions that give rise to considerable amounts of subsidiary products, is less definite in composition than most of the other coloring matters. It had been previously analyzed and found to give a large fraction of coloring matter similar in solubility to Light Green S F Yellowish. Distinguishing these dyes was further complicated by the fact that commercial Light Green S F Yellowish contains small amounts of subsidiary dyes similar to or identical with Guinea Green B. H. L. Lourie states that he considered this differentiation the most diffi- cult of any in connection with the samples. R. W. Balcom used a sys- tematic procedure by which the colors are divided into groups, the solu- q weern veuiny MOTTA UBIO MOTIOR [UOT YStMorex “WS 3437 q weedy vourns MOTOR [UBIO MOT SANG STN q weer vourny q weeny vouIny MOTTIA [TUBPOTA MOTISA [TABIOTA q weer veumy 1q@ Wsee14) BouIny AJ 230819 AMOTTPA [UBIO gq weedy voutnsy 9 adwog H Ped Bet UIsOny uIsOm qaprBog auTye[eg ul OW qojqBog auTyeleg ulso”my yapBog suIyBeg ulsoy YU g nwau0g uIsO qo] Teog auTyRl_eg uIsony U gE NBeou0g uIsOn qo]IBog oulyBI| eq $ ydung q wrurepoyy q umBpoyy UIWIBINy yuIUTBIny q wiuEpoyy ¢ UIMEpOTY UlUIBInNW q ulmepoqy ulmreiny q uluEpoyy uUlUBIny q wluEpoyy ulMBIn Vy q ulUEpoTy uIMeIny $ aduog [Beuryo0y PBS FURIE [Boutyoo0g HES JUNTA r1BoutyIoD yWVY JULIE yo wog FURIE [eourpoog [eouryo0g ow yuRTTIg qoytwog yuEN [AE jeaurqoon AJ esuvig fe eOS OST el [eauryoo09 qoprBog yaRIT[Mg Bouryoog 6 adwuog ‘ap stq} JO UOTJBOYTIUepI uO Yoder puodag , J osuvIig § MOTTAA TOUT Ny YW E Nveou0g JI esuvig J esuvi¢g, § AOTIEA TouIydBN UWE neau0g J esuBIQ U Eg nvoou0g J 8duvig § AoT]9A ToyGdEN UW g Nveou0g yqueiewy J os UBIC_) S MOT]PA TOYWTdEN J esuvig g MOTTPA TOIQYdEN U g Nvau0g [I esuvig J 23uvi9g § MOTTA fouyqdeN UW g nvsou0g J esuvIC S MoT] TOUIGIEN YU gE Nesvou0g ] adwogy ‘SHOLVUAdOOO AM GALUOdAU SUALLYN ONIUOTOO — ‘uuay, ‘a [TAYSe Ny “WOOT “MU “AON WOR MON ‘aunoyT “T 'H “YOTIN “F1OIZ9CT ‘soudvy ‘MH "AON ‘410X MON ‘TsuOTAee “yO ‘O ' “WoySUTYysE MA query “Ha ‘eq ‘Brydyeprytyd ‘uoJuIg “9 ‘OD ‘2 ' SWOUSUTYSE AA ‘10,09 “WW iL ‘Tesue}}9M “A A ‘OMT “OH ‘UL fosBoIqg ‘“WeoATD “MO spsijouy 115 116 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 bilities applied being essentially those given in the tables sent. The details and methods for identification were supplied largely from his own notes. SOLUBILITIES OF COLORS. In any systematic separation of mixtures of the dyeing substances by shaking out their solutions with immiscible solvents, the components will be separated into a number of groups. A knowledge of the solu- bility of colors shows to the analyst at once what must be looked for in any given group. This makes it desirable to have, first, a table showing the dyes arranged in the order of their solubility toward the solvents most important in the analysis; and second, their important chemical properties tabulated in the same way so that the chemist may compare at once the behavior of the different members of a given group. The number of different coloring matters described in any scheme must neces- sarily be limited to a small percentage of the thousands known. B. C. Hesse in 1907 found some seventy coal-tar dyes offered for sale in America for food purposes (Bur. Chem. Bul. 147) and this list has been made the basis of the present table. It includes most of the dyes permitted in France and conforms closely to the group of colors at present sold in Europe for use in food products. In addition, such other dyes have been de- scribed in the table as have actually been found in food products by ana- lysts and finally, a few have been added to represent several important classes of chemical compounds. The solubilities have been determined for the most part with commercial dyes and are therefore given tenta- tively. It will be apparent however, that subsidiary colors of different solubility when present in but small amounts cannot affect the obser- vations greatly as they might do in chemical tests.!_ The aqueous solutions before shaking with the solvents were so made as to contain 0.01 per cent of commercial color. The treatment of the natural colors, most of which are mixtures, is, it must be confessed, very unsatisfactory. 1The data given for amyl acetate was obtained using the ec. p. product of well known manufacturers. Such amyl acetate has since been found to be very variable in solvent power, and analyses made by A. L. Burns in the New York laboratory of a number of samples of the c. p. grade furnished by several of the large manufacturing firms showed in some cases as low as 75 per cent amyl acetate by saponification, the impurity being presumably chiefly amyl alcohol. No considerable amount of free acetic acid was found. Amyl acetate was selected as a solvent because it seemed practically the only one cheap enough to be available and of properties intermediate between those of amyl] alcohol and ether. Mixtures of solvents such as ether and amyl alcohol are open to the objection that when one component is more soluble than the other the washing changes their relative proportion. The amyl alcohol used was Kahlbaum’s ‘‘iso, pyridin free.’’ This doubtless is a mixture of isomers, but experience here has not shown any significant variation in different lots (Author, January, 1915). 19165] ; MATHEWSON: COLORS 117 In the procedure described in last year’s report, any given dye will, in general, appear in several washings obtained in the fractionation. Where the maximum amount comes out, may be judged from the foregoing solubility data. It must be emphasized that the statements made apply only to very dilute solutions. A concentration of 0.01 per cent may be taken as a type. This procedure gives the best results with the acid dyes containing phenolic hydroxyl groups and most of the common dyes used in food belong to this class. With mixture containing only permitted dyes, it affords a fairly rapid means of obtaining the colors in quite pure con- dition in which they may be readily identified. It is more or less unsatis- factory with the lower sulphonated triphenyl methane dyes such as Guinea Green B, and certain others as Methylene Blue, Congo Red, and Naphthol Green B. With some of the basic colors also, the extraction with ether from a strongly alkaline impure mixture is somewhat uncertain, the dyes being sensitive to alkalies and extracted with some difficulty. Where a more careful separation must be made, it is well to begin by treating with one-fourth volume of 25 per cent salt solution, and extract- ing a few times with amyl alcohol. (In all cases in these extractions, any separated solid matter should be considered as belonging to the aqueous layer.) The amyl alcohol is washed with 5 per cent salt solution, then evaporated to dryness on the water bath, the residue moistened with a drop of aleohol and taken up with water. This solution may be made alkaline and the basic dyes removed with ether. Hydrochloric acid is then added until it is about normal and the eosins and lower sulphonated dyes removed with amyl acetate. The acid triphenylmethane dyes are now separated with dichlorhydrin-carbon tetrachlorid mixture. Any color remaining may be separated with amyl alcohol and this added to the amyl alcohol solution of the higher sulphonated colors obtained as follows: Sudans and similar insoluble colors in the residue left on evaporation of the amyl alcohol extract from the sodium chlorid solution will not be taken up by water. Such a residue should be dissolved in ether. In- stead of evaporating the amyl alcohol extract, the dissolved color may in most cases be removed by diluting with gasoline and washing with water (or dilute acid). The salt solution from which the basic and the more soluble acid dyes have been removed is made strongly acid by addition of one-third to one-half volume concentrated hydrochloric acid and extracted a few times with amyl aleohol. The combined amy] alcohol extract is washed once with dilute hydrochloric acid (four to five normal) then several times with fourth-normal hydrochloric acid until the washings contain little coloring matter (the first one or two washings will in all cases contain 118 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 little color because of the acid dissolved in the solvent). The Fast Red and Ponceau dyes may now be removed by washing with water. It may be remarked here that the plan of acidifying strongly, extracting, and then washing the solvent with a more dilute acid is in nearly all cases preferable to gradually increasing the acidity and using a number of portions of solvent. Much more solvent is required in this plan, with a corresponding increase in the proportion of the accompanying impurities and in the difficulties from emulsification. The strongly acid salt solution from which colors soluble in amy] alco- hol have been removed is neutralized with sodium hydroxid or ammonium hydroxid, then made very slightly acid with hydrochloric acid and shaken out a few times with dichlorhydrin to take up strongly sulphonated triphenylmethane dyes. Acid Fuchsin will remain in the aqueous mixture and may be taken up with anilin, or on making strongly acid with phenol. The dichlorhydrin solution is washed with a little 5 per cent salt solution, then diluted with benzol or carbon tetrachlorid and the dye removed with water. When Naphthol Green B is present, the salt solution after extracting the basic dyes cannot be made strongly acid, as this would quickly de- stroy the nitroso color. When its presence is suspected, the neutral salt mixture may be first extracted with dichlorhydrin, washed once with benzol to remove the dissolved solvent, then made half-normal with hydrochloric acid and shaken out with anilin (it is best to add the solvent before the acid). From the anilin solution, the dyes may be fractionated as indicated in the solubility table. The commercial acid yellows (S. & J. Nos. 8 and 9) are mixtures of com- pounds of different degrees of sulphonation. This and their difficult solubility in amyl alcohol renders their separation troublesome, and from _mixtures containing much organic matter they are often best separated by wool. Quinolin Yellow, and in general most dyes made by direct sulphonation of the coloring matter, are likely to yield fractions of different solubilities. The final separation of mixtures of dyes of rather similar solubility will, of course, depend on the selection of some pair of solvents in which they show a decided difference. A fixed plan of analysis beyond a certain point will scarcely be followed as the shade, behavior with acids, and other apparent or easily determined properties of the mixture, will have shown the absence of many colors. Usually a satisfactory fractionation of a pair of dyes can only be effected by the use of two to four portions of solvent and of washing liquid, following a plan similar to that used in a quantita- tive separation.! 1See J. Ind. Eng. Chem., 1913, 6: 26; Bur. Chem. Cir. 113. ope teeveaae gt sta ee Op Ou oes wee base ee Speedie OPO er ees Cre 6b bree ee OPTED LEV ONES e TS Yied-waie oils nail -hereeire Votan svodh ea eS eee , ee ee aa 2 Se Liins be td ne wer vet eee CEES N Ee One PTET eGRTTET Sy tre et vee b eee twee ee ee ed hn eee awed Sse be he PS ese -> -* <¥ 17 a=*s dvlstiginn dL oe csseriqiogrt shotantkes olathe cbojaniee shill wey eduerite s30q tyre) 2 doris oldynl } -betonasnod Le, slredvensea 1A -bejsurtes Uedsae suorlh -baraeraxe [he-dvounlA «betomines Haloop rods, vbedoeas Had sony eet sboteertes ay wy - ob oh Tn 6 As = Kes pre teas we i vb ’ shaders fey raga hs iveepehe tkkerksscks tt ob Vn SR shototrtze Lin deoasid, Nw wieceiapies Annee HE ev - ob > deities sbatnetes Hewloao made eed |. ecco v eek ey Fe eee Ss ees res b -Betontine tedt-ono dradtA , peta Tae wert A (bateedes ¢teq tual 20. dbesbnevcomn Ha seuss. — oe pee EM Aspnes ring wigtel vbapveites He moust ds ob ieee LLL eee ene he) eee CS ee ee eee. Cee seins Ciba Reims Braves erred n sivas poe hee PN DAN Sted SPA Ge FEN y= ee) Cee ee oe eH 64 digi e =~ te) Coe Ee eee eee i) Ceres ee BeBe ois panne en eer Veal s Wy Oe kes inn ope Vereen. | | bbe ee ee eg te ae EXTRACTION OF DYES FROM AQUEOUS SOLUTI Journat or THe Assocrarron or OrriciaL AGRICULTURAL CHEMISTS 9 ONS May 15, Von. 1, No. 1. W. E. Maraewson. (Statements in all cases refer to the amount of dye taken wp vy the _ 4 Ee Se Ee lee ee ee eee ee ee ne 2 ee a a? ae are eg ne ee AMYL ALOOHOL AND DILUTE HYDROCHLORIO ACID DICHLORMYDEIS ASD S7DITM CRLOEIS oT Se 2 ANYL ALQOHDL ANDIO\PER CRN |—————_———— eee CO ee Poe No. RODIN EEHORmOLUON 4. Normal Normal 1/4 Normal 1/16 Normal 1/64 Normal DD pr cnet seb ce Caramel... ... | Little extracted. Little extracted. Little extracted. Little extracted. 4 Little extracted. Little extracted. sets aetna ‘Acid Magenta. - 462 do Larger part not extracted, do P lo aeeavenous eae ‘ ink Chstdasss¥ecnta|> Eee oomnaeedes Anas aréunacssaased ae. Light Green S F yellowish 485 do Little extracted. - e Very little pe Very little ee 6 ery little extracted. Alsseat 21) —, lo ss Saag: = = do do Little extracted. Little extracted. Little extracted. é went ete ew 491 aa do do Larger part not extracted. Larger ne not extracted, Larger part not extracted. ia Bh 440 | Lessthanonebalfextracted. 4. | Onna en nts ueeeenn sneer ert eeteeereectentens|secerererseesteusecececesccersnesteeeacs|scuccecssaeceeuaceneeatensasreriserserss do gee oe 602 | Larger part precipitated. Jews ee eee eeeerneereeeeeee . a 5.4 a iar extracted; atic precipi-] Some extracted; much precipitated. - 6 R.... 108 | None extracted. Larger part extracted. Very little extracted. Sas G s : do do Less than one-half extracted. ‘Little extracted. Acid Yellow R.. 9 GO Pow an cle wawcnvyncceacunssanpers coeds amen|vcosaegGeteccuenbresieteisannaaheessiv sass ‘ Brillisnt Yellow 8... 89 do Larger part extracted. Less than pnetualt eeicactedl! ‘Little extracted. i 692 do do About one-half extracted. do , 399 co Almost al] extracted. do do te as = Ze és 107 do do Larger part extracted. do . Not extracted 106 do do oa ee do 94 do do lo : lo is Naphthol Green B. 398 do Larger part extracted. About one-half extracted. Less than one-half extracted. a ds Azo Carmine Bx 605 do do Larger part extracted. do Little extracted. Orange G..... : uu do -| Almost oJ] extracted. About one-half extracted. Naphthol Black B.. 188 do do do Chromotrope 2 R 20 do do Larger part extracted. About one-half extracted. Less than one-half extracted. Azo Fuchsin G.. 93 do do Almost all extracted. More than one-half extracted. do Water Blue... 480 do do Larger part extracted. do do Less than one-half extracted. Palatine Scarlet 53 do -| Larger part extracted. About one-half extracted. do Ponceau 2 R.. 55 do do ‘do ia Ponceans Be. 56 do -] Almost all extracted. Larger part extracted. a Naphbthol Yellow 8. A Little extracted, 5 do do Less than one-half extracted. Larger part extracted. Less toss cos soncted. do do do do do do a do do About one-half extracted. 103 do do do do 105 do : do do d 604 | Little extracted; much precipitated. 1| Targanpartestrioted a Le ot! Rosindulin 2 G. 606 | Almost all extracted. Acid Violet 4 R. 507 | Less than one-half extracted. Almost all oa Almost all sist About eae extracted. Fast Brown... --| 139 do Larger extracted. i Resorcin Yellow. ween 84 aD: 2 Larger part extracted. do Nearly all precipitated. do Crocein Scarlet 8 B. -| 169 | About one-halt extracted. ied . a do do Little extracted. ) New Red L.......-. 163 | About one-half extracted and precip-|__ = do do Larger part extracted. ead itated. 0 do do Almost all precipitated. do Cotton Scarlet... 146 | About one-half extracted. ma 5 2 Tee oe 4 lo lo part =) Moretliacionehalf‘extracted... 9]. ee, | ee OER ag See ee A an eae erenre tiaeeeesereereneeneessersescssesliceeenee Biswgeemceas sss CCE ES OE RSE eB, SRE PO ERR Ras CERES Ore Sey. Ree peer oe eee ee ee .- Almost all extracted. te Almost all extracted. Almost ‘all extracted: Larger part extracted. Larger part extracted. Larger part extracted it More than one-half extracted; some F ge ES @ do Some extracted; moc precipétated. precipitated. v2 0. lo lo Almost all extracted. Larger part extracted. Some extracted; much precipitated. = do -| Almost all extracted and precipitated.| Almost all extracted and precipitated. do Some extracted; much precipitated. | Some extracted; och stecipitated. Almost all extracted. do do do do a do do do do _do do do do do do Less than one-half extracted. Almost allexigucted. © 6) SO ee I a ee do do do do Larger part extracted. do do FROM AQUEOUS SOLUTIONS BY IMMISCIBLE SOLVENTS. Il cases refer to the amount of dye taken up by the organic solvent.) SILON AND ACID (ANILIN HYDROCHLOMID) 5 PER CENT SODIUM CHLORID SoLtTi0N AMYE ALCOHOL AND 5 PER CENT x (NORMAL) soprgM CARBONATE SOLUTION §-NORMAL SULPHURIC ACID AND AMYL ALCOHOL-GASOLINE MIXTURE (BQUAL voLUMES) DICHLORAYDIIN AND BODIUM CHLOMID BOLUTION 5 per cent sodium chlorid 20 per cent sodium chlorid ee .| Little extracted. do 12 Ae ont all extracted. \ do catnateieed .| Little extracted, Asap R NTF EG e Obra eens OV eSa EM Ue agent hee , "More (hat on extracted, Anis atone? .| Larger part not extracted (three. |. quarters). do extracted. .| Larger part oxtracted (five-sixths), do do do do AMYL ACETATE AND DILUTE HYUROCHLONIO ACID ETHER AND DILUTE HYDROCHLORIC ACID 4 Normal Normal 1/04 Normal 4 Normal Normal 1/64 Normal little extrnotod, Nono oxtrncted, | Very little extracted, penabee Geet? bane +" Semana | '° 2250021 . do do do he Sas ; sarettur do do do bas ; inabla bios canta AMYL ALCOHOL AND NORMAL ‘ SODIUM HYDROXID © part extracted, ED NG A Pee RL POWER eve whee Laks cls dexevs pans on sh pncsiaras dsb ten MA ME dsr ols Ke vep voavae) es an cavereraonecdlossensvonsbg scevovaskon svcoasWuvas rave ddnrohavapatreianne -esevioncesvaiesedsuvas[hnrert¥aneyme¥eRMh -++++cepaceoneie vers y extrnelod, hs i Ces ONL ett iccs isaac huodbbecisplturashokeddespednseil lami revabvensedanechesred rordbesnavannclfta oeciGbsalacenesoartsepesaceocassinaee than ono-halt oxtracted, Ce On Ua, ee CCEA a LUMI 5 Date Say. a Nc assalsl yes acses tevansitivetes (elobartblercncbersenccdovessaioptl Mbceors oclWobeedseosevesesaspoveveseons y extracted, do than ono-hall oxtracted, CS yw a eR a a ee ANF Raeet oe Mw 1 eid cpnecercerentiekvevevaccerevecdecsvale e oxtrautod, Pounsessane canes Pret) oer .| Very little extracted. Se aera || Aa as Co ARTA Ned Ue LALA GAC TR OY OA CMCC Ae EAE, cr dph ys AL aR VUES S Lowy, cuy sacs acaac asl vena Ptants eases osasavd ushsiWEheet|o>lassarstdevsrcocoseceseeucsoneadsraiclecscaccconnensscesacccushectbcccasesgHalvedtsrserestediloah scccssstecessvencee tatoi, little sere Vory little extrmoted, do » extractod, Little extracted, N tmoted. y ull prooipitated. do | apa extmnotod, do F part oxtraobod, do it oll precipitated, RIOR SSE 5 oa 22a AE DIOR Sa SSR aa aa ee a RR IR Gs DERM NNR ROI Oo, limatel ROMS CS rk Se or ii (aaa rai Ocean RN OR (i © part oxtraobed, do 1 part oxtimoted, ‘ do wit all extracted, reer] i ) oxtrnotod; much procipitated, (lo Be Larger purt nob wimovd. ..| Altnowt all extracted. do do do ‘do do do COLOR sss | Caramel! 462 | Acid Magenta 485 | Light Green S FP yellowish 484 | Light Green § F bluish 489 | Cyanole oxtra 491 | Wool Groen 8 440 | Patent Blue 602 | Nigrosin 108 | Ponceau 6 R 8 | Avid Yellow G 9 | Avid Yollow R 80 | Brilliant Yellow S 092 | Indigo Carmine 800) | Curcumin 107 | Amaranth 100 | New Covoin M4 | Tartrasin 808 | Naphthol Greon B 005) | Aso Carmine Bx MM] Orange G 188 | Naphthol Black B 20 | Chromotrope 2 R 98 | Aso Puchsin G 480 | Water Bluo 63 | Palatine Scarlet 65 | Ponoonu 2 R 60 | Ponoonu SR 4 | Naphthol Yellow 8 02 | Palutine Rod 4 | Cryatal Ponoonu 06 | Bordeaux B 103 | Avo Rubin 106 | Mant Red Th 604 | Ano Carmine Gx 600 | Kosindulin 2G 607 | Avid Violot 4 fe 180 | Post Brown BA | Resoroin Yollow 100 | Crovein Searlot 8 1B 108 | Now Kod L 140 | Cotton Searlot 667 | Quinolin Yellow 147 | Vast Brown 267 | Avo Blue 640 | Alizorin Red 8 76 | Drike «| Miri G 46 | Orange T 86 | Orange IT 64 | Brilliant Orange R 14 | Crocoin Orange 97 | Orange Tt 329 | Chrywophenin 95 | Metanil Yellow 68 | Orange LV 101 92 102 510 483 Vast Brown Azo Vlavin Vast Red A Pluorewein Rowolie Acid Meta Chrome Orange Chryeamin G Chrysamin R Sudan Be e8s Orange R. Chrysophenin. Fast Brown. Axo Flavin Fast Red A. Fluorescein... Erythrosin G Victoria Yellow Picric Acid... Erythrowin... Phioxin. Rowe Bengal. Kouln 10 B.. vivaey Martius Yollow...... Aurantli.....ccccccererreeeneree Tose Bengal, Balfron... Versian Berries Quereitron Bark Fumtlo,.. + Logwood Coohineal Arobil...,. Gamboge.....« ‘Turmerio,. ADDAMO. coe Congo Hed? .. Formy] Violet 4 8. Acid Violet N.....+ Methyl Alkali Blue Guinea Green Ih, Brilliant Milling Green 1 Thilo Flavin T. Mothylone Blue Moldola’s Blue... ... vous Wehodamin Byiciccceccceeereeeeeeees Bafranin,...... Ihodamin $1 Magonta,..... : Malachite Green 1, Malachite Green G....... Mothyl Violet, . Oryatal Violet Trimmin G,, Auramin O., Auramin G.. peseever New Mothylone Blue. Biomarok Brown......... Bismarck Brown It Chrysoidin Y, Rhodamin G., Buttor Yellow Anilin Yellow o-Amino-aso-toluone. Bensonearo-cenaphthylamin , Yellow A B, 4 None extracted in ether and sodium hydroxtd solutions, * Almost all extracted to dichlorhydrin and sodtum chlorid solutions pa 3 SSSSss Lew than one-half extracted. Almost all extracted. do do do do Larger part extracted. do do do More than one-half extracted. Larger part extracted, do do do do Almont all extracted. do More than one-ball extracted Lows than one-half extracted Larger part extracted, Almont all extracted. do Very little extracted. Almost all extracted and precipitated. do do do Almont oll extracted. do do do do (8 and 20 por cent sodium More than one-half extracted. About one-hal extracted. Very little ee tractede Little extracted. More than one-half extract. About one-half extracted, do ee. , , 4 ee aa F ‘ F weeks More than one-half extracted, : é < 5 ‘. , ae F Almost all extracted. J Fescct do Almost all extracted. Littlo extracted. Dre stak dus ean py jk ARS Pe ee Larger part oxtracted and precipitated, sdnaont cau 3° ie seeeed. sedeece : 2 ave ies iss is .......-] Larger part extracted. ein 7 . Se Pia es : . . .| Almost all extracted. ae asap aa : do Precipitated. Procipitated. Ler erry Precipitated, Very Uttle extracted. Very little extracted. Little extracted. Almost all extracted. Almost all extracted, None extracted. do Little extracted. Larger part extracted. do do do All extracted. All extracted. All extracted. All extracted. All extracted. .| None extracted, but precipita About one-hall extracted, About one-half extracted. About one-half extracted. More than one-half extracted. More than one-half extracted. Nono extracted. Larger part extracted, Larger part extracted. Almost all extracted. Almost all extracted. Almost all extracted. do 1/64 Normal Normal 1/4 Normal Little extracted. Less than one-half extracted, About one-half extracted. Very little extracted. Larger part not extracted, do Larger part extracted. Little extracted. " Ay A cy Larger part bors Larger part es. do : Loess than one-half extracted, Loss than one-half extracted. mn lo ger do Little extracted. TAttlo extracted. a ° See ae . ‘| Almost al Lee More than one-half extracted, Lous than one-half extracted, Little extrac tected About one-halt extracted, Less than one-half extracted. Very little extracted. Little extracted. Larger part extracted. Lows than one-hall extracted Larger part extracted. 4 Largor part extracted, Almost all extracted. Almoat all extracted. Almost all pei Almost all extracted. lo do do do do do 0 do (5 | SUFiKe 2p --.. | Erika G 85 | Orange I -| Larger part extracted. Little extracted. Almost all extracted. do 86 | Orange IT do do 54 | Brilliant Orange R do do 13. | Crocein Orange do do 97 | Orange R He do 7 do 329 | Chryzophenin do Larger part extracted. Very little extracted. Less than one-half extracted. Little extracted. do 95 | Metanil Yellow do do do do do do 88 | Orange IV do do Larger part extracted. Larger part extracted. Little extracted. do 101 | Fast Brown ce do do do do do 92 | Azo Flavin do do do do do 102 | Fast Red A do Almost all extracted. Very little extracted. do Almost all extracted. 510. | Fluorescein do do do Little extracted. do 483 | Rosolic Acid AMYL ALCOHOL-GASOLINE MIXTURE (3:1) AND NORMAL sopIUM HYDROXID .| All extracted. All extracted. All extracted. Little extracted. Not extracted, 26 | Meta Chrome Orango Almost all extracted. Almost all extracted. Almost all extracted. Precipitated. Precipitated. 220 | Chrysamin G do do do do do 269 | Chrysamin R do All extracted. All extracted. Less than one-half extracted. Little extracted. 10 | Sudan G do Almost all extracted, Almost all extracted. do Very little extracted. 512 | Eosin do do do Larger part not extracted. do 615 | Saffrosin do do do Less than one-half extracted. do 516 | Erythrosin G do do do About one-half extracted. Less than one-bilf extracted. 2 | Victoria Yellow ss sie rae Beer cergramas ates -Paarine|> sun eos -tacstwessbites- ++i da-ceactaree.|’ Aboutone-half extracted” Ral Gans avechemeawants evecs« a [Pamuncnesihisey's seater say » okce 1 | Picric Acid Almost all extracted. Almost all extracted. Almost all extracted. Larger part extracted. Less than one-half extracted, 517 | E do : do do do do 518 | Phloxin do do do do About one-half extracted. 520 | Rose Bengal do do do Almost all extracted. More than one-half extracted. 521 | Eosin 10 B do do do do Almost all extracted. 8 | Martius Yellow do do eat 2 SRE RRS Aace Pegi e ME prod een o--ndoe 4 Gate eee Co eae 4 6 | Aurantia do do do Almost all extracted. §23 | Rose Bengal Very little extracted. Very little extracted. Very little extracted. Very little extracted. Very little extracted. . | Saffron Little extracted, Little extracted. do do PO reg aie || ann AS seas acer he Muse atc cesses | Ors Nilen Werk aBr es cusces svc egchcc call (MEE eee, Mee “ 700 | Persian Berries -| More than one-half extracted. More than one-halr extracted. Larger part not extracted. Larger part not extracted. byt LSPS COB eae HS SR. cen tr Saha cctetey cab x eber | PELcar PES ES (ok ; 090 | Quercitron Bark -| Almost all extracted. Almost all extracted. Almost all extracted. Almost all extracted. Almost all extracted. 698 | Pustic do Larger part extracted. Very little extracted. Very little extracted. Very little extracted. 702 | Logwood Little extracted. Little extracted. do do do 706 | Cochineal Larger part extracted and precipitated.| Larger part extracted and some pre- | Little extracted, Little extracted. Little extracted. 710 | Archil cipitated. Largor part extracted. Larger part extracted. Larger part extracted, Larger part extracted. Larger part extracted. ... | Gamboge Almost all extracted, Almost all extracted. Almost all extracted. Almost all extracted. Almost all extracted. 707 | Turmeric : do do do do do 700 | Annatto Ereipitaied we Mes MO cS cterehiares 240 | Congo Red? Nono extracted. Almost all extracted, 468 | Pormyl Violet 8, B do 464 | Acid Violet N None extracted, but precipitated, 476 | Methyl Alkali Blue None extracted, 4 433 | Guinea Green B Ta oe face te Fh Rc eens eee ee | ne ak ae SLR ae, 438. | Brilliant Milling Gree ETHER AND NORMAL 1/16 Normal ACETIC ACID 1/64 Norma) 1/256 Normal a ae seaSaneseannstataad Bs ..| Larger part extracted. do Almost all extracted. i Gevtatarew ees wt. Meiites a. coesVawssce 658 | Thio Playin T 650 | Methylene Blue 639 | Meldola's Blue 406 Rhodamin S 584 | Satranin 605 | Rhodamin 3 B 448 | Magenta 427 | Malachite Green B 428 | Malachite Green G than one-half extracted. le extracted. than one-half extracted, ut one-half extracted, Little extracted. Little extracted. Leas than one-half extracted. do PCA PE eee oe .. ae ea 451 | Methyl Violet erpartextracted. = | Tmo vet ney args. Bl -exeheewaiseacteads 8s cn tas cement alll iene oral Vea rc, eg .....] 452 | Crystal Violet ii Larger part extracted, About one-half extracted. ae 4 te 490 | Iriamin G ost ull extracted. Almost all extracted, Almost a! I ikrsictad we ae CMe mon oc. cath ot an, scenic Fray Cats patall extracted, bed all” mecsavie Steen eee eee Cece! Perera Perce rrr ery Cee eee errr! bors 426 Auramin G sired Snow al eee New Methylene Blue > Bismarck Brown ae go Bismarck Brown R ab = | Chrysoidin Y Larger part not extracted. 18 | Chrysoidin R EN Whos see Rhodamin B F +e oN Percertr 4 a aaa Rhodamin G See oe Sue - : All extracted. 16 | Butter Yellow a Anilin Yellow o-Amino-azo-toluene Benzene-azo-a@-naphthylamin Yellow A B Almost all extracted. '| Almost all extracted. Almost all extracted. aly «.. | Yellow OB extra Bea th a yp ein SP Oa chemi Pee error tr ee eee) as : ... | @-Amino-azo-naphthalin ee Ba arose aL a - . . “| | Quinolin Yellow S. L io Benzene-azo-a-naphthol B-naphthalenc-azo-a-naphthol Sudan Brown Sudan I Sudan IT Carminaph Garnet Pb cite aw terecns's aeons = Sudan II 1915] MATHEWSON: COLORS 119 Mixtures of dyes of practically identical solubility can in most cases be separated satisfactorily by chemical means or by precipitation reactions. As the fractionation will have removed all except the few dyes belonging to a known group, suitable chemical methods may usually be chosen without difficulty. Since both basic and acid triphenylmethane colors tend to undergo slow intermolecular changes when treated with acids and alkalies (adjust- ment of equilibrium between carbinol, imid and ammonium forms) their complete separation by means of solvents is less simple than that of most other classes. It will be noticed that ether, amyl acetate and amyl alcohol show corresponding properties as solvents except that the colors are more soluble in amy] acetate than in ether and very much more soluble in amyl alcohol than in the others. Dichlorhydrin is intermediate between amyl alcohol and amyl acetate as regards the sulphonated oxy-azo dyes. For the triphenylmethane colors—it is a much more powerful solvent than amyl alcohol. REACTIONS OF INDIVIDUAL COLORING MATTERS. The chemical tests preferred by different analysts will not be the same, but unquestionably those whose chemistry is well understood and that may be considered reactions of certain groups, will be found most useful. Two tests of this type adapted for use with the small amounts of dyes obtained from food products and perhaps not very generally applied at present are those with bromin-hydrazin sulphate and with ammonium cyanid and they have been studied in some detail by the associate referee. Further, in connection with the use of nitrous acid in identifying small amounts of amino compounds hydrazin sulphate has been found to be a suitable reagent for the removal of the excess of nitrous acid, making the test in nearly all cases more convenient and reliable. The diazo compound of Safranin, however, reacts with hydrazin sulphate, apparently. Most widely employed of the common tests for coloring matters are perhaps the color changes produced by reagents on the dyed fiber. A special effort has been made to gather together the best data concerning these. Loomis, in Bureau of Chemistry Circular 63, gives the reactions of a large number of coloring matters when dyed on wool, or for the oil soluble colors, on silk. These observations were made with dyes of estab- lished identity and in this laboratory the statements in the literature have also been checked for some 35 colors with preparations made from pure materials. These general tests, namely, the reactions of the dyes with acids and alkalies, with bromin and hydrazin sulphate, with nitrous acid and 120 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMIsTs [Vol. I, No. 1 finally, the more easily carried out of the well-known reactions involving reduction, have been collected in a set of tables, the dyes being arranged in the order of their solubilities. The properties of the different mem- bers of a group with similar solubilities may thus be at once compared and suitable reactions for their separation selected. These tables because of their length are not included in this report, but it is hoped that they may be brought before the collaborators next year for further elaboration and criticism and perhaps may serve as a nucleus for a logical, coherent chapter on food colors to be incorporated in the official methods. RECOMMENDATION. It is recommended by the associate referee that work bearing on the separation and identification of the common food coloring matters be continued. REPORT ON FRUIT PRODUCTS. By H. C. Gore, Associate Referee. The work planned by the associate referee on fruit products for 1913 consisted in developing further certain studies on the effect of uranyl acetate on the polarization of malic and tartaric acids, and in the appli- cation of a procedure due to Yoder for the estimation of malic and tar- taric acids in various fruit products. In the last study it was expected that eodperative work would be used to determine the value of the method. After the studies on the effect of uranyl acetate had been made, however, the effect of another reagent, ammonium molybdate, on the polarizations of the two acids was studied, and the effect of the respective reagents on the polarization of lactic acid. Time has been lacking for the trials planned of the Yoder procedure. This, therefore, remains as a subject for further work. EFFECT OF VARYING AMOUNTS OF URANYL ACETATE ON THE POLARIZATIONS OF MALIC AND TARTARIC ACIDS. The effect of increasing amounts of uranyl acetate on the polarization of malic and tartaric acids respectively, both free and neutralized, when the concentration of the organic acid was at 1 gram in 100 ce. of solution as polarized, was shown in the 1912 report on fruit products. The study has been carried further, working with more dilute solutions. The data are given in the Table 1. Merck’s ec. p. malice acid and Mallinckrodt’s c. p. tartaric acid were used. All the solutions of malic and tartaric acid were freshly prepared and read in the polariscope at 20°C. The readings were made in 200 mm. 1915} GORE: FRUIT PRODUCTS 121 tubes, using white light from an incandescent carbon filament. The sample of uranyl acetate used was the same used in the studies reported in the 1912 report. It contained 55.475 per cent of uranium. Each gram of malic or tartaric acid thus required 3.2090 grams and 2.8665 grams respectively for one molecular equivalent. Where large amounts of uranyl acetate were present the optical dis- persion was marked. The necessity of using the bichromate cell to ob- tain more correct readings was not realized and the readings are therefore slightly less certain than if the cell had been used. The concentration stated is that of malic or tartaric acid in the solution polarized. TABLE 1. Effect of uranyl acetate on the polarization of malic and tartaric acids. URANYL ACETATE USED FREE ACID NEUTRALIZED ACID ACID AND CONCENTRATION a Specific ; Specific lecular | Grams per olariza- lariza- ceeaianes 100 eat 2 tion ecae roo 3 eben eiVin 1% Sis Via Malic acid: 0.1 gram in 100 ce..... t 0.080 | — 0.30)| — 52 | — 0.65} —113 4 0.160 — 0.90} —156 | — 1.40] —248 2 0.241 — 1.45 —252 — 2.10 —364 1 0.321 — 1.90} -—330 | — 2.80} —486 1} 0.401 — 2.30] —399 — 2.80] —486 14 0.481 — 2.55} —442 | — 2.80] —486 12 0.562 | — 2.75 —477 — 2.70] —468 2 0.641 — 2.75 —477 | — 2.70 —468 rs Hd - ze =e — 2.60 —451 1.6 eae = — 2.60 —451 0.2 gram in 100 ce..... 4 0.160 — 0.85 — 74 — 1.35 —117 B 0.321 — 1.95} —169 | —2.95| —256 2 0.481 — 2.95| —256 — 4.45] —386 1 0.642 | — 3.80} -—330 | — 5.80 —503 1 0.802 | — a0 —408 | — 5.90} —512 14 0.962 | — 5.85 | —464 | — 5.70| —494 12 1.1238 | — 5.65| -—490 | — 5.60] —486 2 1.284 | — 5.65] -—490 | — 5.50] —477 v4 — 5. = — 5.40] —468 0.5 gram in 100 ce..... ry 0.401 | — 2.30] — 80 | — 3.90} —135 4 0.802 | — 4.70} -—163 | — 7.45] —258 3 1.203 — 7.10} —246 | —11.30] —392 1 1.605 — 9.50 | —330 | —14.70] —510 1i 2.006 | —11.70} —406 | —14.55 | —505 1 .404 | —18.45 | -—467 | —14.40 | —500 12 808 —14.30 | -—496 | —14.25| —494 2 209 —14.20 —493 —14.00 —486 3 814 | —13.85 —481 —13.75 | —477 Tartaric acid: 0.1 gram in 100 ce..... i 0.55 95 0.90 156 3 143 1.10 191 1.55 269 3 Ali 1.40 243 2.20 382 1 286 1.70 295 2.50 434 ile 358 2.20 382 2.50 43 122 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 TABLE 1—Continued. URANYL ACETATE USED FREE ACID NEUTRALIZED ACID ACID AND CONCENTRATION : Soecifi ; Specie Molecular rams per | Polariza- Pecie peck Paar on eer aa peat Foerie ee 45 Ven oVi °y. Tartaric acid—Con.: 0.1 gram in 100cce..... iy 0.4380 2.40 416 2.15 373 13 0.502 2.45 425 2.00 347 2 0.573 2.20 382 1.85 321 oF 1.003 1.30 226 0.95 165 5 1.483 (?)0.9) 156 0.75 130 0.2 gram in 100 ce..... rs 0.143 1.15 100 1.85 160 3 0.286 1.95 169 3.10 269 2 0.430 2.80 243 4.35 377 1 0.563 3.45 299 5.30 460 17 0.717 4.30 373 4.90 425 13 0.860 4.65 403 4.60 399 13 1.003 4.85 421 4.20 364 2 1.146 4.55 395 3.85 334 3F 2.007 2.85 247 2.35 204 5 2.866 1.95 169 1.80 156 0.5 gram in 100 ce..... 4 0.358 2.10 73 4.50 156 3 0.717 4.20 146 7.65 265 4 1.075 6.35 220 10.65 370 1 1.433 8.40 291 12.90 448 13 1.791 10.10 250 12.45 432 13 2.150 11.40 395 11.65 404 12 2.508 11.90 413 10.95 380 2 2.866 11.20 389 10.30 357 3 4.300 8.75 304 8.10 281 As in the work reported in 1912 it is here found that when neutralized malic or tartaric acid was employed, maximum values were reached in the presence of smaller amounts of uranyl acetate and the polarizations were higher, than when the free acids were used. Excessive amounts of uranyl acetate caused marked depressions of the polarizations. As before, such depressions were most marked in solutions containing tartaric acid. They were especially large at low concentrations of tartaric acid; for example, the specific rotatory power gradually changed from 434 to 130 in a solution containing 0.1 gram in 100 ce. of neutralized tartaric acid, when the uranyl salt was increased from 1 to 5 molecular equivalents. In the work reported in 1912 it was found that in case of malic acid, concentration 1 gram in 100 cc., depressions in readings caused by ex- cessive amounts of uranyl acetate could be overcome by adding suitable amounts of acetic acid, the fact having been first observed by Yoder. With tartaric acid such depressions could be but partly overcome in this way. ‘The observations recorded in the table below show that the same facts hold at lower concentrations of the respective acids. 1915] GORE: FRUIT PRODUCTS 123 TABLE 2. Effect of acetic acid with uranyl acetate on the polarization of malic and tartaric acids. POLARIZATIONS ACETIC ACID PRESENT Neutralized malic acid Neutralized tartaric acid cc. 0.0 —5.35 +230 0.1 —5.45 42.55 0.2 —5.35 +3 .00 0.4 —5.45 +3 .50 1.0 —5.80 +4.40 2.0 —5.85 +4 50 3.0 —5.65 44.45 4.0 : —5.35 +4.30 The polarizations were made of solutions containing in addition to neu- tralized malic or tartaric acid, 33 molecular equivalents of uranyl acetate, and amounts of acetic acid ranging from 0 to 4 ce. per 100 ce. of solution. The concentration of each solution polarized was 0.2 gram per 100 cc. of malic or tartaric acid. As the polarizations of solutions containing the same amounts of neu- tralized malic or tartaric acid, and sufficient uranyl acetate (1+ and 1 molecular equivalent respectively) to give maximum readings are—5.90 (malic) and + 5.30 (tartaric), the maximum polarization of the solution of malic is shown to be almost entirely restored (from — 5.35 up to — 5.85) by addition of correct amounts of acetic acid, while the polarization of the solution of tartaric acid was but partially restored (from + 2.30 to + 4.50). EFFECT OF AMMONIUM MOLYBDATE ON THE POLARIZATION OF MALIC AND TARTARIC ACIDS. A limitation of the uranyl acetate method proposed by Dunbar and Bacon! for the estimation of malic acid is that it fails in the presence of tartaric acid. Yoder? pointed out that if the sum of the two acids, malic and tartaric, is known, the respective amounts of each may be calculated from the increases in polarization upon addition of uranyl acetate. In a natural product, however, the sum of the two acids is never known with exactness owing to the presence of other acids, acid salts, phosphates and amphoteric amido bodies. Dunbar? has proposed as a measure of the two acids present the amount of oxalic acid formed upon oxidation with alkaline permanganate of the mixture of acids first separated as lead salts. 1 Bur. Chem. Cir. 76. 2 J. Ind. Eng. Chem., 1911, 3: 563; Zis. Nakr. Genussm., 1911, 22: 329. 3 Bur. Chem. Cir. 105 and 106. 124 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMIsTs [Vol. I, No. 1 The method while of promise under certain conditions is not applicable when citric acid or some other substance is present which is precipitated as lead salt and gives oxalic acid on oxidation. If another equation re- lating to the quantities of malic and tartaric acid present could be found to supplement the equation proposed by Yoder expressing the combined polarizations with uranyl salts, it seemed probable that both acids could be determined by optical methods alone; the literature was therefore searched for an account of another element which would cause changes in the optical rotation of malic and tartaric acids. Of the many salts which are described in the literature as affecting the optical rotation of these acids, ammonium molybdate appeared to be of most promise. Gernez! published an extended series of papers dealing with the effect of salts of molybdic and tungstie acids on the polarization of malic and tartaric acids, and on certain optically-active higher alcohols. For malic acid and ammonium molybdate Gernez found a maximum posi- tive reading of (A), = + 740. This is in marked contrast to the maxi- mum reading of (A), = — 501 found by Yoder? for uranyl acetate and malic acid,—a difference of 1241 degrees. Richardson and Gregory* used ammonium molybdate in the study of the estimation of tartaric acid and tartrates. The relation between the polarizations and amounts of tartaric acid present was found not to be linear, but the polarizations very rapidly increased until at the higher concentrations a specific rotatory power of 861.8° was reached. Rimbach and Schneider‘ studied the effect of ammonium molybdate and other salts of molybdenum, also tungsten, uranyl, titanium, zirco- nium, and other elements on the polarization of quinic acid. The ex- tremely high polarizations caused by ammonium molybdate with malic and tartaric acids were not found in their study. Rosenheim and Itzig® studied the compounds of beryllium with malic and tartaric acids, and with substituted derivatives of succinic acid. Crystalline d-beryllium tartrates were prepared of high specific rotatory powers. Upon dilution remarkable stability was shown by the fact that the high specific rotatory powers persisted. Tentative structural formulae were suggested. Itzig® later studied the effects of salts of molybdenum and other metals on the polarization of tartaric acid and many of its salts and upon malic acid, without, however, having in mind the estimation of the re- spective acids. Data similar in many respects to facts described by Gernez were obtained. 1 Gernez papers in Compt. rend., 1887 to 1890 incl. 2 Loc. cit. 3 J. Soc. Chem. Ind., 1903, 22: 405. 4 Zits. physikal. Chem., 1903, 44: 467. 5 Ber. d. Chem. Ges., 1899, 32: 3424; 1900, 38: 707. ® Ibid, 1901, 34: 1372, 2391, 3822. 1915] GORE: FRUIT PRODUCTS 125 The interesting inversions in the polarizations found by Gernez were studied by Grossman and Pétter,! and by Maderna? as well as other phases of the subject. With the exception of the paper by Richardson and Gregory, in which malic acid was not considered, however, no data has been secured with a view to the development of a method for the estimation of the two acids by use of ammonium molybdate. PRELIMINARY STUDY. The ammonium molybdate used contained 54.595 per cent of molyb- denum. The calculated percentage of molybdenum is 57.72 for the salt of the formula (NH,)sMo;O2,. The deficiency noted was probably due to the presence of water of crystallization. The analysis was made by igniting MoO; according to directions given by Treadwell-Hall (Quantita- tive Analysis, Vol. II, p. 250). Each gram of malic and tartaric acid re- quired 1.3123 and 1.1717 grams of the ammonium molybdate respectively for one molecular equivalent of respective acid to one of molybdenum. On account of the high dispersive powers of solutions of ammonium molybdate, the bichromate cell was used throughout. Asin the work with uranyl acetate all solutions were made up and read at 20°C. Tubes of 200 mm. length, and white light from an incandescent carbon filament were used. The effect of increasing amounts of ammonium molybdate on polariza- tions of solutions of free and neutralized malic and tartaric acids respec- tively is shown in the following tables. Each solution contained as polar- ized 1 gram in 100 cc. of one of the acids either free or neutralized. TABLE 3. Effect of ammonium molybdate on the polarization of malic and tartaric acids. AMMONIUM MOLYBDATE USED POLARIZATIONS ACID AND CONCENTRATION Suet ene bee Free Neutralized CG SVE Malic acid: {egramvanyl OOlcemneestierc 0 0.000 fee sane yo 0.132 — 1.30 — 1.10 i 0.328 — 0.65 — 1.55 3 0.656 + 4.95 — 2.10 2 0.984 +12.20 — 2.05 1 1.312 —19.35 — 0.95 1i 1.640 +21.35 + 1.30 14 1.968 +21.40 + 4.15 13 2.296 +21.10 + 6.40 2 2.624 +20 .55 + 7.80 33 4.594 +1905 +10.50 5 6.562 +17 .80 +10.55 Ibid, 1905, 38: 3877; Zits. physikal chem., 1906, 56: 577. * Aitt. R. Accad. dei Lincei, Roma, 1910 (5) 19 (2): 130. 126 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 TABLE 3.—Continued. AMMONIUM MOLYBDATE USED POLARIZATIONS ACID AND CONCENTRATION aureaienta Mees tone Free Neutralized °V. SV. Tartaric acid: WerameinyOOlCesn-seeis-re ee 0 0.000 0.95 2.20 ay 0.118 3.60 3.70 3 0.292 7.20 6.15 a 0.586 12.05 9.80 3 0.878 15.85 13.40 1 1.172 19.80 17.25 11 1.464 23.75 21.15 13 1.758 27.30 24.75 12 2.050 30.05 27.45 2 2.344 31.70 28.15 3} 4.100 36.95 29.75 5 5.856 36.20 30.05 Perhaps the most conspicuous feature of the above table is the peculiar inversion of the curves for free and neutralized malic acid discovered by Gernez. Neutralized malic acid solution was made far less highly active than the solution of the free acid, and the solution of neutralized tartaric acid was made less active than that of free tartaric acid, although the difference is less striking. The questions arise—can these rotations be stimulated by acetic acid, and can conditions be found under which a linear relation exists between concentration of malic or tartaric acid and optical activity? EFFECT OF ACETIC ACID ON POLARIZATIONS OF SOLUTIONS OF AMMONIUM MOLYBDATE AND NEUTRALIZED MALIC OR TARTARIC ACIDS RESPECTIVELY. Each of the solutions polarized contained 1 gram in 100 ce. of neutral- ized malic and tartaric acids respectively, and from 1 to 33 molecular equivalents of ammonium molybdate. The results are given in Tables 4 and 5. TABLE 4. Effect of acetic acid on the polarization of solutions of ammonium molybdate and neutralized malic acid (1 gram in 100 cc). POLARIZATIONS WITH AMMONIUM MOLYBDATE USED AEE) a) 1 molecular 14 molecular 2 molecular 3 molecular equivalent (1.312 | equivalents (1.968 | equivalents (2.624 | equivalents (3.936 grams in 100cc.) | gramsin 100,cc.) | grams in 100 ce.) grams in 100 cc.) cc. tn 100 ce. SLE, ie cis CMG 0.0 — 0.50 + 4.55 + 8.00 +10.20 0.2 + 3.15 +10.90 +13 .95 +15 .55 0.4 + 8.15 +17 .95 +26 .65 +23 .60 1.0 +10 .20 +26 .55 +38 .05 +37 .35 2.0 +10.50 +26 .90 +41 .65 +42 .30 4.0 +10.70 +27 .30 +4260 +43 .25 6.0 +10.65 +27 .05 +42.70 +43 .50 10.0 +10 .45 +27 .20 +42 .70 +483 .50 16.0 +10 .45 +26 .95 42.50.” fh) eee : 1915] GORE: FRUIT PRODUCTS 127 TABLE 5. Effect of acetic acid on the polarization of solutions of ammonium molybdate and neutralized tartaric acid (1 gram in 100 cc.) POLARIZATIONS WITH AMMONIUM MOLYBDATE USED Pisroutie! Dien) af) 2 molecular 2} molecular 24 molecular 34 molecular equivalents (2.344 | equivalents (2.636 | equivalents (2.928 | equivalents (4.100 grams in 100 cc.) grams in 100 ce.) grams in 100 cc.) grams in 100 cc.) ce. in 100 cc. Vie ois Sve Ve 0.0 28.10 28 .50 29 .00 29.50 0.2 29.45 29.75 30.05 30.30 0.4 30.85 30.90 31.10 31.15 1.0 32.45 33.30 33.20 33.30 2.0 35.20 35.25 35.15 35.10 4.0 36.40 36.20 36.65 37 .00 6.0 37 .05 37 .35 37.35 38 .05 10.0 37 .50 37.75 38.25 38.70 16.0 37 .65 38 .20 Beer eo mec me peace In case of malic acid it was expected that maximum readings would be obtained when from 1 to 1} molecular equivalents of ammonium molybdate had been added and sufficient acetic acid. Instead the readings increased very rapidly with increasing ammonium molybdate, approaching appar- ently maximum readings when 3 equivalents of the molybdenum salt was present and acetic acid in amounts ranging from 10 to 16 grams in 100 cc. The results observed in case of tartaric acid are even more unexpected. Regular increases with no indication of approach to maximum readings occurred as ammonium molybdate and acetic acid increased. SEARCH FOR CONDITIONS IN WHICH THE POLARIZATIONS OF NEUTRALIZED MALIC AND TARTARIC ACIDS MADE ACTIVE BY THE PRESENCE OF AMMONIUM MOLYBDATE AND ACETIC ACID ARE PROPORTIONAL TO THE AMOUNTS OF THE RESPECTIVE ACIDS PRESENT. Two series of solutions of neutralized malic and tartaric acids con- taining 2 grams in 100 ce. of the respective acids were prepared, each solution containing 3.5 molecular equivalents of ammonium molybdate. Each series consisted of six solutions in which the amounts of acetic acid present varied from 0 to 20 ce. in 100 ce. Portions of each solution were diluted to concentrations of 2, 1, 0.5, 0.4, and 0.2 gram of optically active acid in 100 cc., and readings were made on all solutions in the polariscope. The readings are given in Table 6. 128 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMIsTS [Vol. J, No. 1 TABLE 6. Polarizations of fixed solutions of neutralized malic and tartaric acids with ammonium molybdate and acetic acid. NEUTRALIZED MALIC ACID NEUTRALIZED TARTARIC ACID Sa ACETIC ACID PRESENT : = SEE Ay Polarization Ree Polarization Se grams tn 100 cc. cc. tn 100 cc. Vis wiVe is “Vi 2.0 0.0 10.55 92 60.70 526 2.0 4.0 85.25 739 71.70 622 2.0 6.0 86.25 748 74.80 649 2.0 8.0 86.90 754 77.00 668 2.0 10.0 87.25 757 78.85 684 2.0 20.0 87.00 754 86.00 746 1.0 0.0 10.80 187 29.55 513 120 2.0 42.20 732 35.30 612 1.0 3.0 42.90 744 36.20 628 1.0 4.0 43.10 748 36.90 640 1.0 5.0 43 .30 751 37 .40 649 1.0 10.0 43.35 752 38 .45 667 0.5 0.0 8.20 284 14.20 493 0.5 1.0 20.55 713 16.50 572 0.5 1.5 21.05 730 16.75 581 0.5 2.0 21.25 737 16.75 581 0.5 2.5 21.20 735 16.75 581 0.5 5.0 21.30 739 16.60 576 0.4 0.0 7.25 314 11.00 477 0.4 0.8 16.25 705 12.85 557 0.4 122 16.75 726 12.85 557 0.4 1.6 16.80 729 13.05 566 0.4 2.0 16.85 731 13.00 564 0.4 4.0 16.95 735 12.80 555 0.2 0.0 4.15 360 5.25 455 0.2 0.4 7.80 676 5.50 477 0.2 0.6 8.10 702 5.35 464 0.2 0.8 8.15 707 5.35 464 0.2 1.0 8.05 698 5.35 464 0.2 2.0 8.40 728 5.15 446 The results show that the rotation of malic acid stimulated to or near the maximum point by the presence of suitable amounts of ammonium molyb- date and acetic acid, are practically proportional to the concentration of malic acid. With tartaric acid this is not the case, but the specific rota- tory power falls with increasing concentration indicating dissociation with dilution of the highly rotating complex of molybdenum and tartaric acid. It was hoped that conditions of strictly linear relations would exist between the increases in polarization and amounts of respective acids present. If this were the case the estimation of malic and tartaric acids in the pres- ence of each other could be easily made by determining the increases in polarization after the addition of uranyl acetate and ammonium molybdate respectively. As, however, such relations were not found, it remains for tables to be constructed from data yet to be worked out from which may be read the amounts of malic or tartaric acid corresponding to a deter- 1915] GORE: FRUIT PRODUCTS 129 mined change in optical rotation. No method for the estimation of the two acids in the presence of each other by polarimetric means alone is in sight at present. STUDY OF THE EFFECT OF URANYL ACETATE AND AMMONIUM MOLYBDATE ON THE POLARIZATION OF LACTIC ACID. In the report of 1912 it was shown that the presence of sucrose and invert sugar did not interfere perceptibly with increases in optical ac- tivity when uranyl acetate was added to solutions containing malic acid. Two substances, however, exist widely distributed in nature, one of which, mannite, is known to be made very active in optical rotation by uranyl salt (Grossman)! and by ammonium molybdate (Gernez)? and the other, lactic acid, might, it was thought, interfere as it is optically active and possesses the carboxy grouping present in both malic and tartaric acids. Tests were therefore made to determine the extent to which polarizations of lactic acid are made active, by uranyl acetate and ammonium molybdate. It may be stated here, however, that as by precipitation as barium salt by the Yoder procedure the sample is freed from possible presence of mannite or lactic acid and studies of the effects of the reagents on the polarizations of these substances are without interest where this method is used. They are of importance however, in connection with short methods like that of Dunbar and Bacon. The sample of high grade commercial lactic acid used polarized directly at 20°C. in a 200 mm. tube at 20°C. at + 16.4°V. Titration showed that the sample contained 88.2 per cent of total lactic acid, of which about 20 per cent was present as lactid. For the polarimetric study 20 grams of the sirupy acid was diluted, treated with a slight excess of solutions of sodium hydroxid, heated to boiling, and boiled for a few minutes. It was then acidified with hydrochloric acid, then made just alkaline, and the boiling continued. The pink color of the phenolphthalein remained un- changed in intensity. It was therefore concluded that all the lactic had been converted into the sodium salt of lactic acid. The solution was now cooled, exactly neutralized and made up to 100 ce. The solution polarized at + 3.00°V. = (A), = + 2.95. Of this solution portions of 5 cc. were treated with slightly more than one molecular equivalent of uranyl acetate and ammonium molybdate respec- tively, diluted to 100 ce. and polarized. The polarizations were as follows: Wranyllacetaterns cee are coon oon eae cee —0.6°V., (A)p = —11.8 Ammonium molypdatesmemecccsteetiacccr cscs A +0.5°V., (A)> = + 9.8 1Zts. Ver. d. Zucker-Ind., n.f., 1905, 42: 1058. 2 Loc. cit. 130 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 Thus, though uranyl acetate and lactic acid polarized to the left, and ammonium molybdate and lactic acid polarized to the right, the changes are quite small. Further tests showed that these polarizations were not increased by adding acetic acid. SUMMARY. Studies were made on the effect of increasing amounts of uranyl acetate on the polarizations of malic and tartaric acids respectively when the acids were present in concentrations of 0.1, 0.2, and 0.5 gram per 100 ce. As previously found with concentrations of 1 gram in 100 cc. excessive amounts of the reagent caused depressions in optical activity particularly marked in case of tartaric acid. By adding suitable amounts of acetic acid the maximum activity was almost entirely restored in case of malic acid, but only partly restored in case of tartaric. Studies on the effect of ammonium molybdate on the optical rotation of malic and tartaric acids showed the known great magnification of the rotations, and showed great differences in optical activity depending on whether the particular acid was free or neutralized. Acetic acid added to solutions containing neutralized malic or tartaric acid greatly increased the optical activities. With malic acid, in the presence of suitable amounts of ammonium molybdate and acetic acid, very large positive polarizations were obtained which were practically proportional to the amounts of malic acid present. With tartaric acid very large positive polarizations were obtained in the same manner, but the readings were not proportional to the amounts of tartaric acid present. The polarizations of lactic acid were found to be so slightly affected by either uranyl acetate or ammonium molybdate that it could hardly inter- fere, if present, in solutions being polarized. RECOMMENDATIONS. It is recommended that the study of the effects of uranyl acetate and ammonium molybdate respectively, on the polarizations of malic and tar- tarie acids, in the presence of known amounts of acetic acid, be continued, with the view of constructing tables from which the amounts of malice or tartaric acid present can be read, when the change in optical rotation due to either reagent is known. : That the Yoder procedure for the estimation of malic and tartaric acids be studied in connection with such tables. Two papers giving the results of preliminary work on “Determination of Total Solids in Fruit Juices and on Fruit Jellies and their Manufacture” were read by L. W. Andrews. 1915] HARTMANN: WINE 131 REPORT ON WINE. By B. G. Harrmann, Associate Referee. This year’s work on wine was devoted to the determination of tartaric acid. As was pointed out in the Proceedings of 1912, the customary method of determining tartaric acid, as described in Bulletin 107, Re- vised, does not give satisfactory results when applied to wines containing considerable amounts of free tartaric acid. It was decided, therefore, to try out the modification of this method as outlined by Hartmann and Koff (Bur. Chem. Bul. 162, p. 71) with a view of ascertaining how this method compares with the bulletin method, and of determining its relia- bility in the hands of different analysts. Accordingly, two California white wines were sent to six collaborators. No 1 was a straight wine and No. 2 was the same wine to which 300 mg. of tartaric acid per 100 ee. were added. DIRECTIONS TO COLLABORATORS. 1. Mix the contents of the quart and pint bottles of each set and filter. 2. Transfer 20 cc. of the filtered wine to a 400 cc. beaker and bring to an incipient boil to drive out carbonic acid. Add 20 ce. of recently-boiled distilled water and titrate with tenth-normal sodium hydroxid solution using phenolphthalein as indicator. DETERMINATION OF TARTARIC ACID. Bulletin method.—Follow the directions given on page 86 of Bulletin 107, Revised, using phenolphthalein as an inside indicator instead of litmus paper in the titra- tion of tartaric acid. Hartmann and Eoff method modified.—Measure 50 cc. of filtered wine into a 250 ce. beaker and completely neutralize with normal sodium hydroxid. (The amount of sodium hydroxid necessary to accomplish this is found by multiplying by 0.25 the number of cubic centimeters of tenth-normal sodium hydroxid required to neutralize the 20 cc. of wine as determined under 2). To the wine so treated add water to make 100 ec.; add the equivalent of tartaric acid corresponding to the number of cubic centimeters of normal sodium hydroxid added. To find this amount of tartaric acid multiply the number of cubic centimeters of normal sodium hydroxid added by 0.075, which will give the grams of tartaric acid required to convert all the tar- tarie acid present in the wine into bitartrates. Weigh exactly this amount and dissolve in the 50 ce. of neutralized and diluted wine. Prepare the tartaric acid to be used by pulverizing pure acid and drying for 2 hours in a water oven. Record the weight of tartaric acid added, and after it has gone into solution add 15 grams of potassium chlorid and 2 ce. of glacial acetic acid and stir until the salt is completely dissolved. Add 20 cc. of 95 per cent alcohol and stir one minute; set aside for 15 hours and collect the crop of cream of tartar crystals as directed in the bulletin method and titrate using phenolphthalein instead of litmus paper as indicator. CALCULATION OF TOTAL TARTARIC ACID. Bulletin method.—Proceed as described in Bulletin 107, Revised; that is, add 1.5 ec. to burette reading and multiply by 0.015. 132 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 Hartmann and Eoff method modified —Add 1.5 ec. to burette reading and multi- ply by 0.015. From this amount subtract the tartaric acid added and multiply the difference by 2. CALCULATION OF FREE TARTARIC ACID AND CREAM OF TARTAR. In order to figure the free tartaric acid and the cream of tartar, it is necessary to determine the alkalinity of the ash. Alkalinity of ash.—Measure into a platinum dish 50 ce. of the wine at 20°C., evap- orate on a water bath to a sirupy consistency and ash carefully at a dull red heat, never allowing the fumes to ignite. If difficulty is experienced in burning the ash white, leach with hot water. To the white ash add about 5 cc. of water and 3 drops of a 10 per cent solution of ammonium carbonate; evaporate and heat below redness to expel ammonium salts. Add hot water to the ash and transfer the whole to a 7 cm. filter with successive small quantities of hot water until the filtrate amounts to 50 cc. This gives the portion of the ash soluble in water; transfer the portion insoluble in water remaining on the filter to a platinum dish and ash. To the residue add 10 ec. of tenth-normal sulphuric acid and about 10 ec. of water and bring to a boil. Break up the insoluble portion with a stirring rod and again bring to a boil; titrate back with tenth-normal alkali. Subtract the amount of alkali required from the 10 cc. of acid used and multiply the remainder by 2 to find the alkalinity of the water-insoluble portion. Use phenolphthalein as indicator. To the portion soluble in water add 10 cc. of tenth-normal sulphuric acid and bring to a boil. Titrate back with tenth-normal alkali using phenolphthalein as indicator. Subtract the number of cubic centimeters required to titrate back from 10 and multiply by 2 to find the alkalinity of the water-soluble portion. The total alkalinity of the ash is the sum of the alkalinity of the water-insoluble portion and the alkalinity of the water-soluble portion. The following terms may be used in the calculations: (a) The acidity of the tartaric acid contained in 100 cc. of wine expressed in terms of tenth-normal acid. (b) The total alkalinity of the ash expressed in terms of tenth-normal acid. (c) The alkalinity of the water-soluble ash expressed in terms of tenth-normal acid. (d) The alkalinity of the water-insoluble ash expressed in terms of tenth-normal acid. To find (a), divide the total tartaric acid content in 100 cc. of wine by 0.015. With the help of these terms (a), (b), (¢), (d), the tartaric acid in the free state, and as cream of tartar, may be figured in the following manner: (1) If (a) is greater than (b): Cream of tartar = 0.0188 X (ce), Free tartaric acid = 0.015 X [(a) — (b)], Tartaric acid to alkaline earths = 0.015 X (d). (2) If (a) equals (b) or smaller than (b); but greater than (c): Cream of tartar = 0.0188 X (ce), Free tartaric acid = 0, Tartaric acid to alkaline earths = 0.015 X [(a) — (c)]. (3) If (a) is smaller than (c): Cream of tartar = 0.0188 X (a), Free tartaric acid = 0, Tartaric acid to alkaline earths = 0. Example: Total tartaric acid = 0.680 grams per 100 cc. 1915] HARTMANN: WINE 133 Total alkalinity = 31.0 ce. of tenth-normal acid. Water-soluble alkalinity = 24.6 ce. of tenth-normal acid. Water-insoluble alkalinity = 6.4 cc. of tenth-normal acid. (@) = aS = 45.3 cc. of tenth-normal acid. (b) = 31.0 ec. of tenth-normal acid. (c) = 24.6 cc. of tenth-normal acid. (d) = 6.4 ec. of tenth-normal acid. In this case (a) is greater than (b), therefore, 24.6 * 0.0188 = 0.46 grams per 100 ce. of cream of tartar; (45.3 — 31.0) 0.015 = 0.21 gram per 100 ce. of free tar- taric acid; 6.4 X 0.015 = 0.10 gram per 100 cc. of tartaric acid to alkaline earths. To be assured that the tartaric acid added in the Hartmann and Eoff modified method is pure, it is advisable to titrate 0.1500 gram of the dried acid with tenth- normal sodium hydroxid using phenolphthalein as an inside indicator. This should require 20 cc. of the soda. In reporting your results please enclose this titer on the 0.1500 gram of tartaric acid. RESULTS OF COOPERATIVE WORK. The following tables give the results obtained by four of the collabora- tors: Coéperative work on tartaric acid. TOTAL TARTARIC ACID FREE TARTARIC ACID CREARMGOR | |S e ane eaten aia DEAD ORGS LES) TSSEN ee Bulletin Hartmann | TARTAR Bulletin Hartmann method | 22d Eoff method | 22d Eoff method method gram gram gram gram gram per 100 cc. | per 100 cc. | per 100 cc. | per 100 cc. | per 100 cc. SAMPLE 1: H. L. Lourie, New York, N. Y..| 0.17 0.18 Sete eee ERAS J.R. Eoff, Jr., Washington, D.C.| 0.17 0.17 0.05 0.05 0.05 M. J. Ingle, Washington, D.C..} 0.18 0.19 0.05 0.07 0.08 H. C. Fuller, Washington, D.C.| 0.19 0.20 0.04 0.08 0.09 AVELALES cars sicieios sete esists anette 0.18 0.19 0.05 0.07 0.07 SAMPLE 2: H. L. Lourie, New York, N. Y.. 0.43 0.45 J. R. Eoff, Jr., Washington, D.C.| 0.41 0.45 0.06 0.29 0.34 M. J. Ingle, Washington, D.C..| 0.45 0.48 0.06 0.33 0.37 H. C. Fuller, Washington, D.C.| 0.45 0.49 0.04 0.33 0.37 IAN CLAD Os Ss cc Ame cacios cows, hoe 0.44 0.47 0.05 0.32 0.36 Results on alkalinity of ash. (ec. tenth-normal acid per 100 cc. wine). SAMPLE 1 SAMPLE 2 ANALYST SS Se el nh | So ee Se a ee Soluble Insoluble Soluble Insoluble alkalinity alkalinity alkalinity alkalinity J. R. Eoff, Jr., Washington, D.C... Pat 5.0 3.0 4.4 M. J. Ingle, Washington, D. C...... 2.8 5.0 3.0 4.8 H.C. Fuller, Washington, D.C..... 2.2 5.0 2.3 5.7 PAV CLAP Cs cts tennant nals Mclelareietiecer 2.6 5.0 2.8 5.0 134 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 It is evident from the figures in the tables submitted that the determina- tion of tartaric acid as described in Bulletin 107, Revised, gives results materially too low and that the proposed method gives results which are very close to the truth. The average content of total tartaric acid in the straight wine No. 1, determined by both bulletin and proposed methods is 180 and 190 mg., or 185 for the average of both methods. Accordingly, wine No. 2, which is wine No. 1 to which 300 mg. of tartaric acid were added, should show 485 mg. The following are the results obtained by the methods under discussion: Results on total tartaric acid. TOTAL TARTARIC ACID ANALYST Bulletin method | Hartmant and Eoft mg. per 100 cc. mg. per 100 cc. edb oonrie, sNe way OL INe vere este 430 450 JR. Hoff, Ir:, Washington, Di@..-.22...-.- 410 450 M. J. Ingle, Washington, D. G............... 450 480 He G]uller Washington). Grease. oe 450 490 INSVET AGES T Se. chit ee rcinireCroet ee iaet esar trey 435 468 From these determinations, it will be seen that, although the results of the several analysts are not concordant, the bulletin method consistently gives results which are lower than those obtained by the proposed method and that the bulletin method out of a total of 485 yields only 435, and proposed method 468 mg. of tartaric acid. NOTES BY THE COLLABORATORS. M. J. Ingle: In all ash work it was found essential to burn off in an electric muffle at a ‘‘dead heat” or one that could not be distinguished by sight, in order to prevent fusing. In the tartaric acid determination by the method outlined in Bulletin 107, Revised, the time of stirring was lengthened to about 2 minutes, as in cases where small amounts of tartaric acid were present it was found necessary to positively start the formation of the precipitate. Im drying the tartaric acid used in the modified method, a Petri dish was found to be a convenient and useful container as the powdered dried acid will keep dry for days in it. It was suggested to try out the addition of a neutral tartrate instead of using the free acid as used in the proposed method outlined by Hartmann and Eoff, the original intent being to have the neutral tartrate take care of all free tartaric acid. In order that a salt available to most chemists might be used, Rochelle salt was chosen. As this salt contains 4 mole- 1915] HARTMANN: WINE 135 cules of water of crystallization, drying in an oven was found to be im- practicable. Drying in a desiccator was next tried, but even in this mild dehydrating atmosphere, the salt was found to lose 12 per cent of its weight in a short period of 80 hours. It was decided to use the pure salt as received and run a blank with each series of determinations. Two blank determinations on the same weight of salt were tried by the bulletin method with the omission of potassium acetate; to one enough free tartaric acid was added to just form the acid salt; in the second the same object was accomplished, depending upon the 2 cc. of acetic acid to form the acid salt. Both procedures yielded excellent results, giving nearly per- fect yields of tartaric acid. The fact that only 53.17 per cent of the weight used is tartaric acid, makes accurate additions of tartaric acid easier by this means. A wine containing 0.20 gram per 100 cc. of free tartaric acid in addition to about 0.22 of combined acid was used for the experiment. The wine was titrated for total acidity and the factor for determining the addition of Rochelle salt was calculated to be 0.1409 times the number of cubic centimeters of normal alkali found to be neces- sary for complete neutrality. It developed in the course of these determinations that the recovery of tartaric acid was not complete in the determinations on wine when the acidity was not first neutralized, so that the experiment did not prove successful in doing away with the neutralizing, but was shown to be of material benefit as suggesting Rochelle salt as a stable and efficient means of introducing weighed amounts of tartaric acid into the completely neutralized wine or grape juice. In the following determinations the bulletin method was used, omitting the addition of potassium acetate in all but B. In A and B the wine was completely neutralized, C was not neutralized. Determinations I and IT are the blanks to which reference has been made. Determination of tartaric acid using Rochelle salt. TOTAL ORIGINAL SUBSTANCE DETERMINATION Seeaoee ee RGU RCO FABDaSraaom gram gram gram gram per 100 cc. per 100 cc. per 100 cc. per 100 ce. \iiih\ee Re oe crear A 0.793 0.842 0.838 0.416 Wine B 0.793 0.842 0.836 0.414 Wamnedeeies venice C 0.793 0.842 0.800 0.378 Solutionl.....-...- I 0.793 0.845 0.841 0.422 Solutionvee s+ -..- II 0.793 0.422 Os428i Wi” Rats The relative efficiency of the Hartmann and Koff modification and Kling’s rocemate method (Annales des Falsifications, 1910, p. 239; 1911, p. 185) was tried out in the Charlottesville laboratory. Two wines rep- 136 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, Nu. 1 resenting the extremes in the ratio of free to combined tartaric acid were used. Sample 2096 contained 0.31 gram per 100 ce. of free tartaric acid out of a total of 0.53; the second sample (3262) had a total of 0.34 of which only 0.05 was free. The principle of the Kling method is the for- mation of calcium rocemate in a favorable medium for a very complete precipitation. This is accomplished by adding lime and laevo tartrate of ammonia in suitable quantities and titrating the calcium rocemate with permanganate. The time factor is in favor of this method, but owing to the expense of the left tartrate and the fact that the tartrate solu- tions are not stable, the Hartmann and Eoff method appears to be the more practicable for most laboratories, especially those not possessing a microscope. The following table is self-explanatory; 25 ec. of wine were used in each determination: Comparison of the Kling and the Hartmann and Eoff methods for total tartaric acid. TOTAL TARTARIC ACID SAMPLE NO. TESTS nee Hartmann and Eoff Kling method mnethoal gram per 100 cc. gram per 100 cc. 1 0.559 2 0.554 3 0.547 4 0.554 Bure AVOTARE s:cj5\s)sjc;cicieus 0.553 1 0.342 2 0.344 3 0.347 4 0.337 3262 Average. 2.2 .c0<05 0.342 J.R. Eoff, jr.: It was found that a very simple modification of the usual method for tartaric acid determination or of the recently proposed method of Hartmann and Eoff will enable an analyst in most cases to determine the total tartaric acid content of a wine or grape juice in less than one hour’s time. The major portion of the time ordinarily necessary for a tartaric acid determination is the 15 or more hours that the solution is allowed to stand to accomplish complete precipitation. This period may be re- duced to 15 to 30 minutes, depending upon the amount of tartaric acid present, by the use of a mechanical stirrer. The following table will illustrate the practicability of SLSR the stirrer for the longer method of precipitation: 1915] HARTMANN: WINE 137 Determination of tartaric acid using a mechanical stirrer. HARTMANN AND|HARTMANN AND TARTARIC ACID |HARTMANN AND| EOFF METHOD; | EOFF METHOD; Bose nena CONTENT EOFF METHOD STIRRED 30 STIRRED 15 MINUTES MINUTES gram per 100 cc. | gram per 100 cc. | gram per 100 cc. | gram per 100 cc. Tartaric acid solution.......... O48) OH Bis ae Bree ORAS 2 eat ei do OFS ieee OPARS elle tl onan do OE2Z4E Di) MAPeee rere ee wea terey 0.238 do 932: Re a MERA ac ste ty 1 | a seats 0.236 do O24 ee Taree tases 0.238 do ORLZO LE Seen cone: 0.110 do 0.048 0.042 0025 0an ene do 0.048 0.042 1OLOL20 olen Moe 2 California Hock wine........... 20.254 0.240 Hoos } Besuates 9 California Sauterne............ 20/221 0.215 { eee } ae California Zinfandel............ 20.233 0.211 {0.248 } Site . : f 0.100 Caiioatin SGeaysabcsoqssescau) I. dndae 0.088 1o.099 |f co Crltornrareortictn ce cccshiecaceelh leer 0.113 Oe 22 Se aa eee California Sherry No. 2........ 0.101 10.089 } qnerne @alifomiayPortyNon2-eacsecce-||) ceeae 0.120 OM20i9 ||" tees 1 Stirred 60 minutes. 2 Content determined by bulletin method. As can be seen it is possible to determine accurately as little as 0.1 gram per 100 cc. of tartaric acid in less than an hour by the use of the Hartmann and Eoff modification of the bulletin method if the mechanical stirrer is used in conjunction therewith. For amounts of tartaric acid materially less than 0.1 gram per 100 cc., it is not advisable to make use of the shortening of the time, since even with an hour’s stirring the pre- cipitation is incomplete. The stirrmg apparatus is so well known that a description would be superfluous. The speed of stirring used in the above tests was as high as could be applied without loss of solution by spattering. It is suggested that, in the absence of a stirring apparatus, a bottle containing bits of glass rod or glass beads would answer the same purpose if the solution under examination were placed therein and vig- orously shaken during a similar period. The application of the stirrer points out additional advantage of the Hartmann and Eoff method, that of furnishing quickly the tartaric acid content of low acid wines. NOTES BY THE REFEREE, The reports by the four collaborators show that the modified method by Hartmann and Eoff gives very satisfactory results and that the bulletin method is not reliable in cases where much free tartaric acid is present in a wine. Before proposing the method for adoption as a provisional method, more work should be done especially as to its applicability to red wines. 138 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 The report of M. J. Ingle furnishes additional evidence that the pro- posed method gives highly satisfactory results. In his report he shows that determinations of tartaric acid made in two wines by the proposed method and the Kling rocemate method give very close checks. A very interesting contribution to the method of tartaric acid determi- nation is made in this report, wherein a neutral salt of tartaric acid is substituted in place of the tartaric acid addition, as provided in the Hart- mann and Eoff method. This modification of the proposed method should be more thoroughly investigated, because it may prove to be a more desirable procedure than the proposed one. As regards the report by J. R. Eoff, jr., it should be said that the method is a valuable contribution because it makes quick work a possi- bility without forfeiting accuracy, in all cases where amounts of tartaric acid greater than 0.10 gram per 100 cc. are present in the liquid under examination. REPORT ON BEER. By J. G. Ritzy, Associate Referee.’ The work on beer consisted in making a critical study of the provisional method for phosphorie acid (Bur. Chem. Bul. 107, Rev., p. 92), and a comparison of this method with other methods as to accuracy and time of determinations. Two samples of beer, a malt beer marked No. 1, and a beer made from 55 per cent of malt (by weight) and 45 per cent of corn, marked No. 2, were submitted to collaborators to determine phosphoric acid expressed as milligrams of phosphoric acid per 100 cc. by the following methods: METHODS FOR COOPERATIVE STUDY. Method 1.—Uranium acetate method: Bureau of Chemistry Bulletin 107, Revised, page 92, article 15. Method 2.—Measure 25 cc. of the beer, 15.6°C., free from carbon dioxid, evaporate and ash at alow red heat. Transfer with dilute nitric acid to an Erlenmeyer flask and proceed according to Bulletin 107, Revised, page 4, No. 2 (bi), Total Phos- phorie Acid. Method 3.—Measure 25 cc. of the beer, 15.6°C., free from carbon dioxid gas, and add 25 cc. of standard calcium acetate. Prepare calcium acetate as given in Bulle- tin 107, Revised, page 21, 2 (a). EXvaporate and ash and proceed as in Method 2. Method 4.—Wet combustion method: Measure 25 cc. of the beer, 15.6°C., free from carbon dioxid, into a 500 ec. Erlenmeyer flask, and proceed according to Bulle- 1 Read by M. J. Ingle. 1915] RILEY: BEER 139 tin 107, Revised, page 2, Total Phosphoric Acid (as), and then according to page 5, 2 (b1), Total Phosphoric Acid. The collaborators were requested to note also the effect of corn upon the amount of phosphoric acid in the beer. Reports of seven collaborators are tabulated as follows: RESULTS OF COOPERATIVE WORK. Coéperative results on phosphoric acid in beer. (mg. per 100 cc.) PHOSPHORIC ACID COLLABORATOR Method 1 Method 2 Method 3 Method 4 Sample |Sample |Sample |Sample} Sample |Sample| Sample | Sample 1 2 1 2 1 2 1 2 R. W. Hilts, Seattle, Wash. | 90.6) 58.3 | 90.7) 46.0 | 1101.6] 146.9] 290.9 3.3 5 90.6} 53. 91.1) 45.9 | 4101.2 | 147.2) 788.5 101.4] 47.4] 395.3 | *43.9 101.0} 46.9] %84.0 | *44.0 Sicha 4101.2 | 444.9 sparse 4101.3 | 445.6 H. C. Fuller, Washington, 1D) (CAsee Coane aaa rn ar ....| 50.0 | 91.2) 54.4 | 101.2] 48.8] 94.4] 44.0 S. H. Ross, Omaha, Nebr..| 90.7; 46.5 | 100.9] 47.8 | 100.3, 48.3] ..... ae Tale 18% Mead, Philadelphia, Ura auis jeVarse ote cisseleouays e-8 108.1) 48.8 | 58.1) 50.2 | 103.8] 51.4] 48.2] 44.0 100.7| 47.7 54.1) 47 103.0] 52.8] 77.8] 45.6 90.1 62.7) . 104.3 90.4 ee 91.2 59.6] . 102.4 80.2 E. Fettijobn; St. Paul, Minin ey tessa ce nee 101.0) 49.0 | 53.0} 47.0 | 104.0] 47.0] 101.0] 45.0 E. Grab, Nashville, Tenn...) 88:1) 48.5 | 87.9} 45.3] 90.6) 45.5) 91.0) 45.5 S820 P48h6nll eaee line Seen |e otere de AG Riley, Washington, ; WAC Se cae aseh, a p ose 102.4) 51.4 | 42.6) 44.0 99.6} 46.0] 102.4] 48.4 1Unburned carbon particles not burned out. 2 Using nitric acid and hydrochloric acid only. 3 Using nitric acid, hydrochloric acid and sulphuric acid. ‘Using nitric acid and sulphuric acid only. COMMENTS OF ANALYSTS. R. W Hilts: In Method 1 the uranium solution described by Sutton (‘‘Vol. Anal.,’’ 10th ed., p. 309) was used and standardized on a solution of microcosmic salt under the same conditions as in titrating the beer. The microcosmie salt solution was standardized gravimetrically by the magnesia method. The good agreement of duplicates is accidental as the end point could hardly be determined closer than 0.25 ee., which means an experimental error of = 5 mg. of phosphoric acid per 100 ce. of beer. This method gives erroneous results and is wholly unsuited to the esti- mation of such small amounts of phosphoric acid. Method 2 is better, but there appears to be a loss of phosphorus in the all malt beer. Method 8 gives excellent results. In the determinations marked ! the unburned carbon particles were not filtered out, which made the end point a little difficult in 140 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 the titration. In the other cases the nitric acid solutions of the ash were filtered through paper, the paper washed and ashed and the residue taken up in acid and added to the main solution. In Method 4 in determinations marked °, this laboratory used 30 ce. of nitric acid and 5 ce. of hydrochloric acid, boiling down to a small volume and then diluting and proceeding to the precipitation, just as directed (as). This does not destroy all organic matter and the result appears to be low. In determinations marked * the same amounts of acid were used but 8 cc. of sulphuric acid were added and organic matter was totally destroyed by one or two further additions of nitric acid, after evaporation to sulphur trioxid fumes. In determinations marked ‘ the beer was put in a 500 ce. Kjeldahl flask with 30 cc. of nitric acid and 8 cc. of sulphuric acid and was gently boiled down to the trioxid fumes, when all organic matter was destroyed without further additions of acid or any attention. I think this is the most con- venient method of performing the wet combustion. In * and 4, finally 25 to 30 ec. of water were added, the solution boiled for several minutes, and then rinsed into beakers, which were used in all cases instead of flasks, for the molybdate precipita- tions. The phosphomolybdate precipitates were filtered on asbestos and, for washing, water containing a little ammonium nitrate was used. There is then no danger of the precipitate washing through the filter in a colloidal condition, as often happens with plain water. In my opinion, Method 44, destroying all organic matter, is preferable to Methods 1 and 2 in point of accuracy, and to Method 3 in ease of manipulation. H.C. Fuller: It is evident that the presence of corn materially reduces the con- tent of phosphoric acid. Method 1 is of no value with dark beers and ought not to be recommended as a method which might at any time be made the basis of a court procedure. One is not sure of his end point within 5 cc. in either direction. The prevailing opinion in this laboratory is in favor of Method 3. S.H. Ross: In Method 1, the titration factor is too large to allow close determi- nation of the phosphoric acid. In Method 2, by ashing at a low temperature in an electric muffle, the use of calcium acetate prescribed in Method 3 appears unneces- sary. In Method 4, the formation of the yellow precipitate was very slow; after standing 45 minutes and filtering, further precipitation occurred in the filtrate. There appears to be no advantage in the wet combustion method over the direct incineration and the method was abandoned without ascertaining the cause of in- complete precipitation. H. B. Mead: Of the four methods, Method 3 appears the most accurate in con- sistently giving the highest results obtained. The possible reason for this is that the calcium acetate might assist in retaining the organic phosphorus. The provis- sional uranium acetate method apparently indicates about 90 per cent of the total phosphoric, assuming that Method 8 yields the correct figure. For Method 4 would say that results on Sample 2 should probably be thrown out, as the Erlenmeyer flasks containing 25 cc. of sample, 30 ce. of concentrated nitric acid, and 5 ce. of concentrated hydrochloric acid were heated on the steam bath for some hours be- fore digestion over a flame. All portions of Sample 1 for treatment by Method 4 were measured out before Sample 2 was opened, which eliminates the possibility of confusion in marking. No explanation is offered for the low results obtained on Sample 1, Method 4. The analysis proceeded smoothly at all stages. Digestion was made with a low flame. c Method 1 is, of course, the quickest, requiring about 15 minutes after the beer has been freed of the gas and providing the volumetric solution is ready for use. Methods 2 and 8 take a little over one working day as it was hardly possible to do 1915] RILEY: BEER 141 more than evaporate and ash in less than 7 hours. Ashing in Method 2 was more rapid than in 3 as the calcium salts seemed to prevent speedy oxidation by the latter method. In fact, an entirely white ash was obtained in none of the determinations tried by this method, although the ash was moistened repeatedly. Moistening the ash with nitric acid might shorten the time and should give a white ash. Method 4 consumed about 2 hours, the oxidation being complete in about 1 hour. More vigorous boiling might shorten the time, but no information is available concerning the effect of such treatment on the results. The lack of agreement in Method 4 on Sample 1 may be due to inexperience with the method. E. Pettijohn: In conducting the determination according to Method 1, some difficulty was experienced in judging the correct end point of the titration. This was especially the case with Sample 1. It seems desirable in working on beer to make a trial titration according to the procedure described in the official method, then repeat the titration, adding 2 to 3 drops of the standard uranium acetate solu- tion at a time toward the last. The characteristic end-point indication will finally develop and no serious difficulty should be noted. In working with Sample 1 according to Method 2, low results were obtained. The sample was reduced to ash in two different ways, first over an open flame under a cover of platinum foil, then in an electric muffle. The same results were obtained by each procedure. No reason can readily be assigned for the low results on Sample lby thismethod. They are nearly as low 4s results obtained by the same method on Sample 2. In applying Method 4 to Sample 1, it was found necessary to repeat the treat- ment and evaporation with the acid mixture, as the single treatment appeared to be quite inadequate to get the sample in right condition. The results turned out satisfactorily, duplicate determinations agreeing very closely. Of the four methods subjected to trial, Method 1 is preferred, chiefly because it was found possible under ordinary conditions to complete the two determinations inside of a half hour. The time required by any one of the other three methods varied from four to five hours. Method 4 would be second in order of preference. Each of the results reported in the above tabulation is an average of two or three closely-agreeing check determinations. The peculiarity noted in the results ob- tained by Method 2 on Sample 1 is a matter that does not seem to be easily explained. The results, however, are as correct as could be obtained, in this as well as in the other determinations reported in the table. E. Grab: In Method 1, I wish to call attention to the fact that the uranium acetate solution is standardized against a colorless phosphate solution. The method used in this standardization is that given in Sutton’s “‘Volumetric Analysis,” 10th edition, Phosphoric Acid and Phosphates, page 308, paragraph 2, and page 309, paragraph 3. There is a decided color present in the beer, and in titrating this against the uranium acetate, it will be on a different basis than the standardization of the uranium acetate. The amount of uranium acetate used is small in comparison to the amount of beer taken. No investigation has been made to determine whether using a more dilute solution will affect the method. If not, I should judge that a solution about half as strong as given in Sutton would be more accurate. Because of the color present in the beer, determination of the end point requires practice. The two results given by me on both of these beers in this method are fairly close. Consider- ing that the standard solutions are already made, the method is a very quick one, taking less time than the other three methods. Method 2 is similar to the method used in this laboratory for the determination of phosphoric acid in vinegar. I consider it accurate if the heat is kept down when 142 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEmMiIsTS [Vol. I, No. 1 the ash is burned. I wish to call attention to the precipitation of the ammonium phosphomolybdate. If certain conditions are not followed, this precipitate is too fine and passes through the filter. Method 3 is practically identical with Method 2, the only difference so far as I can see being the longer time it takes to evaporate. I fail to see anything gained by the addition of the calcium acetate unless it is the fact that the beer is ashed in less time, and that the chances of losing phosphoric acid by too high heat is less. Method 4 is no more accurate, in my opinion, than either 2 or 3. The process of destroying organic matter by the addition of nitric acid, hydrochloric acid, and heat, brings in several factors which I consider objectionable—the large amount of fumes and the risk of the vessels breaking. I found also that if all organic matter was not destroyed, the results obtained are low, and that it took several additions of nitric acid and hydrochloric acid to destroy all this organic matter. I am of the opinion that if this method is followed, it is best to evaporate the beer in the flask nearly to dryness, then to digest. The time taken, following Method 4, is decidedly longer than that by any of the other methods. Referring to the enclosed sheet which contains the values obtained by these dif- ferent methods, I wish to call attention to the fact that I found that the addition of corn meal lowered the total phosphoric acid 48.3 per cent. In both samples the results obtained with Methods 3 and 4 compare favorably. In Sample 1, the results with Methods 1 and 2are fairly close. In Sample 2, the results obtained with Method 1 are higher than those obtained with any of the other methods. All measurements were made at 20°C. DISCUSSION OF RESULTS. An inspection of the table indicates that Method 3 gives the most concordant results. Although this determination takes from 4 to 5 hours, the actual time of manipulation is not more than 20 minutes. It was checked with the gravimetric method as follows: Duplicates of Samples 1 and 2 were analyzed by F. W. Smithers and E. L. Wilcox, Bureau of Chemistry, according to Method 3. The ash was taken up with water and nitric acid, filtered, and washed; residue burned, and again taken up with dilute nitric acid and added to the main portion. The phosphoric acid was then precipitated as ammonium phosphomolybdate, and finally weighed as magnesium pyrophosphate as follows: Weight of magnesium pyrophosphate 0.0383 and 0.0392 gram in Sample 1, 0.0172 and 0.0190 gram in Sample 2; corresponding to the weight of phosphoric acid in 25 ce., 0.02443 and 0.02500, !0.01097 and 0.01211; which equals milligrams of phosphoric acid in 100 ec., 97.72 and 100.00, ——, and 48.44. These gravimeteric determinations were repeated, using Method 4 for the oxidation of the organic matter, with the following results: Weight of magnesium pyrophosphate 0.0386 and 0.0388 gram in Sample 1, 0.0180 and 0.0178 gram in Sample 2, corresponding to the weight of phosphoric acid 0.02462 and 0.02474, 0.01148 and 0.01135; which equals milligrams of phosphoric acid per 100 cc., 98.48 and 98.96, 45.92 and 45.40. 1 Discarded, loss of precipitate. 1915] ADAMS: DISTILLED SPIRITS 143 From observation of the gravimetric determinations, it is seen that the concordant results of Method 3 agree with the gravimetric results, within experimental error. During the last three years, Method 4 has been used in the Bureau of Chemistry and found to check with the gravimetric method. From the tabulated results, however, it is readily seen that it is a method which re- quires more experience to obtain correct results than Method 3. There- fore, Method 8 is preferable. Method 2 is an uncertain method, with a great danger of loss of phos- phorie acid and should be abandoned. Method 1 appears to be accurate for light beers, but dark beers have to be diluted many times in order to get a proper end point and results are uncertain. Therefore, it is reeommended that Method 3 be adopted, in place of Method 1, as the official method in the determination of phosphoric acid in beers. The effect of corn in lowering the amount of phosphoric acid present in the beer is also to be noted. Your attention is called to corrections which should be made in the dextrin determination as given on page 91 of Bulletin 107, Revised. The phrase ‘‘in the original beer,” in line 6, evidently means in the beer after hydrolyzing, and the final word of the method should be “beer” and not wort. “Multiply the amount of maltose in the original beer by the factor 11.053 and substract this from the total amount of dextrose found after hydrolyzing, and multiply the remainder by 0.9, giving the dextrin in the original beer.” The factor 1.053 takes account of the fact that 19 parts of maltose yield 20 parts of dextrose after hydrolyzing with dilute acid. The factor given in Bulletin 107, Revised, is the reciprocal of the proper factor. For an example of this calculation correctly stated see Koenig, Untersuchung Landwirtschaftlich und Gewerblich Wichtiger Stoffe, 3d Ed., page 665. REPORT ON DISTILLED SPIRITS. By A. B. Apams, Associate Referee. As no work had been planned for the past two years the present asso- elate referee had no unfinished work to take up. Chemists who have done much spirit analyses know that excessive amounts of aldehydes in the spirit, if not removed, are extracted, oxidized, 1 This is on the assumption that the anhydrous maltose table is used. This fac- tor, however, can not be less than 1.000 nor more than 1.053. 144 —s ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. J, No. 1 and titrated, raising the final result higher than is correct for the method used. It was thought that this error would be a good one to take under consideration for 1913. J. M. Doran, of the Bureau of Internal Revenue and the associate referee, did a certain amount of preliminary work, in order to find a basis for collaboration. In this way the amount of work required was reduced to a minimum. Briefly, this preliminary work consisted in adding to two samples of distilled spirits containing aldehydes, varying amounts of m-phenylenedi- amin hydrochlorid, determining the amount of aldehydes left therein, and the effect of these aldehydes on the fusel oil results. It showed that, saponification with sodium hydroxid removed a portion of the aldehydes; when aldehyde content was about 20 parts per 100,000 it removed about three-quarters; and that increasing amounts of this m-phenylenediamin hydrochlorid had to be added as the content of aldehyde increased; 0.2 of a gram seemed to be satisfactory for contents of about 20 parts of aldehyde and 0.5 was found to be sufficient for excessive amounts. Two samples were prepared and sent to those collaborators who had signified their willingness to assist, with the request that the fusel oil determination be made in these samples as given in Bulletin 107, Re- vised, page 98, then the sample be treated by the following method: After the esters are saponified and the spirit distilled, add 0.5 gram of m-pheny- lenediamin hydrochlorid, reflux for one hour, redistill, and proceed as usual. Sample 1 was a so-called brandy called “Grappa,” which is used a great deal in California and which always contains large amounts of aldehydes. The aldehyde content of this sample was 151 parts per 100,000. Sample 2 was made from alcohol, amyl alcohol, and aldehydes. The aldehyde content was 28.7 parts per 100,000. Coéperative results on determination of fusel oil. FUSEL OIL BY ALLEN-MARQUARDT METHOD ANALYST Sample 1 Sample 2 Untreated Treated Untreated Treated W. W. Karnan, Cincinnati, Ohio........| 269.0 109.0 97.0 68.0 J. I. Palmore, Washington, D. C..... .. 278 .6 102.7 108.5 76.5 J. M. Doran, Washington, D. C.........| 348.0 98.3 76.6 61.5 It is not necessary to dwell on these results; they show clearly that when 20 parts of aldehyde are present the result is appreciably affected, 1915] GOODNOW: VINEGAR 145 and that when larger amounts are present the result obtained by the untreated method is hopelessly incorrect. The associate referee has the honor to submit the following recom- mendation, that the Allen-Marquardt method for fusel oil determination be modified as given in Bulletin 107, Revised, page 98 (b) Allen-Marquardt method, fourth line after ‘‘is collected” by adding ‘“‘ Whenever aldehydes are present in excess of 15 parts per 100,000, to distillate add 0.5 gram of m-phenylenediamin hydrochlorid, reflux for one hour, distill 100 cc., add 25 cc. of water and continue distillation until an additional 25 cc. is collected.” It is recommended further that the associate referee be directed to con- fer with the chemists, engage in spirit analysis, do such collaborative work as necessary, and submit at the next meeting a revised Allen-Mar- quardt method containing such change in details and manipulation as experience justifies. REPORT ON VINEGAR. By E. H. Goopnow, Associate Referee. No collaborative work on vinegar has been undertaken this year. The referee work of the association for the past two years has been given over to a general and very complete revision of the methods of analysis, which has resulted in a modification of the official and provisional methods as published in Bulletin 107, Revised, in the application of a number of standard determinations to the analysis of vinegar, and in the develop- ment of new tests which have been found of value in detecting adultera- tion or establishing the purity of a sample. This completed scheme of examination is given in the Proceedings of the association for 1911 (Bur. Chem. Bul. 152, pp. 126-7). These methods of analysis have been submitted to collaborative exami- nation in connection with the referee work. Their reliability has been further established in the very large number of check analyses made in connection with the official inspection work of the Bureau of Chemistry. The recommendations of the referee that several of these methods be fur- ther studied were not carried out, owing to the fact that the present ref- eree was appointed as a substitute for the original referee too late to give opportunity for the necessary preliminary work for the development of a scheme of investigation, which would be of value in studying these methods. It is, therefore, recommended that a study of the methods suggested by the referee for 1912 be included in the collaborative work of the association for the following year. 1 Not read at this meeting. 146 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 REPORT ON FLAVORING EXTRACTS. By A. E. Pau, Associate Referee. The collaborative work this year was confined to two extracts, vanilla and peppermint. In planning the work, the suggestions of last year’s referee were adhered to quite closely. VANILLA EXTRACT. COMPOSITION OF AUTHENTIC VANILLA EXTRACTS. In following the recommendation for a further study of the composition of pure commercial vanilla extracts, it seemed desirable to secure analyses of extracts prepared from the greatest variety of beans, prepared by as many different methods as possible. To do this, each collaborator was requested to either make or obtain elsewhere extracts of absolutely known origin, and report his results, obtained by the official methods. Unfortunately very few responses were received to this request. The results and comments of the collaborators follow: Collaborator’s analyses of authentic samples of vanilla extracts. a Nee aes B. B. WRIGHT H. E. WOODWARD 2 MANN 2 Noss ie = i DETERMINATION = 2 Sa - am =™ Meeicant Palate hee Se isse 3 gn 22 (U.S. P-) beans Ho Ege] 3 | gS | 35 Sle a = ra A B A B Specificlpravity/at15:0/6.0| eeu icles ie cite eeeeterenll ese crave 1.009 T0104), =.=): || eee Vanillin (gram per 100 cc.)| 0.20 | 0.14 0.25 | 0.2128 | 0.2992 0.28 0.23 0.049 0.051 Lead number............. 0.51 | 0.50 0.65 } 0.4727 | 0.6448 0.74 0.76 0.10 0.10 Coloriextract, xedi.-5---2-| eee lbaeee BOLE creeon | Mecca 1175. 1185. 1112. 1110. Color extract, yellow..... 1222. Weaker lesaweesl| Tob taeee Tl eee 2c Ml aie eeceteel leer Color filtrate, red......... 1 Ey PT sooo lesapece 12 16. 4 4 Color filtrate, yellow...... AU) |Wrencono|) Gapsod! koceeacess oesoadae tl Saccooos |! Socs 032 Color in filtrate, red...... Svea | Weert nee nie 6.9 8.7 3.6 3.6 Color in filtrate, yellow. . eal (Po pcdicite| oAaaccnl MOM Sa BOemdoetion Mebcrerccal) pecosco0 Color insoluble in amyl alcohol ( percent)....... scares | isan EO i sop se 4 hanéooc 18. 20. 5.2 5.5 Alcohol (per cent by vol- NING) Peet toksce ie senteeis Apto 4\ acd CEU W oSocoe I, asaaco 53. BES WWepeastaa |} saccouss Ash (grams per 100 cc.) FaRe tciaaoel (heriot< al), sGueded Macadae 0.40 0.40 0.10 0.10 Solids (grams per 100cc.)} .... | .... SiO lh Peer cl] keen ae 23.2 23.6 1.08 0.29 Sugars (grams per 100cc.)} .... | .... ORC a eee lee meyers 21.3 21.8 0.83 0.08 1 Brewer’s scale. The extracts examined by A. W. Hanson were manufactured by Ridenour Baker Grocery Company, Kansas City, Mo. The Mexican extract was prepared from 40 1 Read by Julius Hortvet. 1915] PAUL: FLAVORING EXTRACTS 147 pounds of prime Mexican beans to 50 gallons of finished product with 25 pounds of sugar. The Tahiti and Bourbon was made from 35 pounds of Tahiti, 6 pounds of Bourbon, and 25 pounds of sugar. The ingredients were macerated three months or over in 25 gallons of alcohol and 15 gallons of water, drawn off, diluted to 50 gallons, and allowed to settle. H. J. Wichmann’s samples were manufactured by Brown Brothers Mercantile Company, Denver, Colo., from 12 ounces of Bourbon cuts per gallon of alcohol diluted half with water. These were digested at 110°F. for 24 hours in a machine made by the Hardesty Company. After completion of the process a little sugar was added. B. B. Wright reports that the Mexican beans were imported from Vera Cruz, Mexico. They were high grade and of uniform length, averaging 20cm. In making the extract the only deviation from the U.S. P. was that 3 days instead of 24 hours were consumed for percolating the extract. The Bourbon beans were declared by the shipper to have been grown in Reunion and Madagascar. In making the extract, the percolation was continued for 3 weeks, thus extracting quite completely all valu- able ingredients. The extracts examined by H. E. Woodward were made in the laboratory by C. 8. Brinton from high grade Mexican beans, which had become much dried out and brittle in the laboratory. The original extracts were made according to the U.S. P. .The dregs were macerated and percolated with 60 per cent alcohol to one-half the volume of the original extract. The results on these second extracts should, there- fore, be divided by two in order to place them on the same basis as the original extracts, or for comparison with the results obtained by Winton and Berry (Bur. Chem. Bul. 162). WICHMANN’S QUALITATIVE COUMARIN METHOD. Wichmann’s qualitative coumarin method originally appeared in Bureau of Chemistry Circular 95. Since its publication the author has slightly modified the directions to read as follows: Slightly acidify 25 cc. of the extract, if alkaline, with sulphuric acid, add 25 cc. of water, and distill until yellow decomposition products appear. To the distil- late, containing the vanillin and coumarin, add 15 to 20 drops of 1 to 1 potassium hydroxid, hastily evaporate the distillate to 5 cc., and transfer to a test tube. Heat the test tube over a free flame until the water completely evaporates and the residue fuses to a colorless, or nearly colorless, mass. Coumarin, if present, will be con- verted into potassium salicylate. If decomposition products have been allowed to distill over, the potassium hydroxid solution will become yellow, turning black as the fusion point is reached. In this case the fusion will not become colorless, but usually an appreciable diminution of the black color will be observed. It is well to add more potassium hydroxid when decomposition products are present to prevent the danger of burning the salts toa char. Cool the melt and dissolve in a few cubic centimeters of water. Transfer the solution to a 50 cc. Erlenmeyer flask and acidify slightly with 25 per cent sulphuric acid. The amount of solution should not be over 10 ce. Finally distill the solution into a test tube containing four or five drops of neutral 0.5 per cent ferric sulphate or ferric chlorid. The efficiency of the reagent should be tested by a few cubic centimeters of a solution containing not more than 0.1 mg. of salicylic acid per cubic centimeter. If coumarin is present in the original 148 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 extract, an amethyst or purplish color will develop, the intensity being directly proportional to the amount of coumarin. This method and two samples were submitted to the collaborators. The samples consisted of a true U.S. P. vanilla extract, with and without the addition of 0.01 per cent of coumarin. The collaborators were requested to apply Wichmann’s method, re- porting results as ‘‘positive’’ or “negative,” and to make a statement as to whether, in the routine examination of miscellaneous vanilla extracts, genuine and artificial, it is desirable to make a preliminary test by Wich- mann’s method, followed, in the case of negative results, by the evapora- tion directly of the original ether extract obtained by the official method, or whether it is preferable to proceed in all cases according to the com- plete official method, without the preliminary examination according to Wichmann. The results reported were as follows: Wichmann’s qualitative test for cowmarin. COLLABORATOR ike a Se wrt 0.01 wa cor phe enya Chicago lll eeeer eee naereerre negative positive H. C. Fuller, Washington, D. C.............. do negative A. W. Hanson, Kansas City, Mo............ do positive H. A. Halverson, St..Panl,-Minn....,.4..... do do P. W. Holtzendorf, Memphis, Tenn.......... do do W. B. D. Penniman, Baltimore, Md......... do do S#He Ross; Omaha Nebreeeeecre sees ene do de : rR ) oO Aro Valin tOtta wary Cam sie ses trteeii cit et aes { do do H. E. Woodward, Philadelphia, Pa.......... { ag Benin BaBs Wrights New sorks NiaYonssce cence do negative H. J. Wichmann, Denver, Colo..............- do positive The following comments were made by collaborators as to the desira- bility of applying Wichmann’s test preliminary to the official examination: E. H. Berry: Very little time will be saved by Wichmann’s method. H.C. Fuller: Inadvisable to make preliminary test by Wichmann’s method. A.W. Hanson: Not much advantage in making preliminary test by the Wichmann method. There might be, if combined with Folin’s method. H. A. Halverson: More explicit directions should be given. Preferable to proceed according to the complete official method in all cases. P. W. Holtzendorf: I do not consider that Wichmann’s method affords any saving of time, and I prefer to operate by the official methods without preliminary examination. C. F. Jablonski: With ample time I should prefer the official method to that of Wichmann’s, but the latter is nevertheless very valuable. 1915] PAUL: FLAVORING EXTRACTS ; 149 S. H. Ross: Wichmann’s test is valuable as confirmatory evidence, but does not afford any saving of time. H. E. Woodward: I prefer to use the complete official method. B. B. Wright: When a hasty examination for the presence of coumarin is desired, I consider that Wichmann’s method is very valuable. However, if time permitted, I should prefer the official method. FOLIN’S COLORIMETRIC VANILLIN METHOD. Folin’s colorimetric vanillin method appeared in the Journal of Indus- trial and Engineering Chemistry, volume 4, page 670, and depends upon the color produced in a phosphomolybdic phosphotungstic acid solution, by the addition of alkali. The method appeared very desirable for mix- tures which may contain substances that would interfere with the official method, such as fats and benzoic acid. Unfortunately, however, since sending out the samples it was found that various substances give positive tests with Folin’s reagent. Among these might be mentioned an infusion of cocoa beans, which appears to contain 1 per cent of vanillin when ex- amined by Folin’s method. Salicylic and benzoic acids give a very slight result. Collaborators were requested to determine vanillin by this method on the same two samples which were submitted for Wichmann’s coumarin method. The vanillin contained in the two samples was the same, and was very carefully determined independently by four analysts, using the complete provisional method. The results reported by the collaborators were as follows: : Collaborators’ results on vanillin by Folin’s colorimetric method. VANILLA EXTRACT PURE VANILLA EXTRACT witH 0.01 PER CENT COLLABORATOR COUMARIN ADDED Folin’s Official Folin’s Official method method method method grams grams grams grams per 100 cc. per 100 ce. per 100 cc. per 100 cc. Pept bercy, Chicago Mlvane: seen 0.19 0.19 0.20 0.20 H. C. Fuller, Washington, D. C......... 0.20 hee 0.20 spied A. W. Hanson, Kansas City, Mo: ss for. |» 048 oa 0.18 Rees H. A. Halverson, St. Paul, Minn....... 0.20 ate 0.21 re Paw: Holtzendorf, Memphis, Merny Ol 22 a: 0.21 ee C. F. Jablonski, New York, INE XGA. 0.19 See 0.19 dee A. E. Paul, Chicago, aa seein rca. ae 0.20 is 0.20 W. B..D: Penniman, Baltimore, Md.... 0.20 one 0.21 Eee S. H. Ross, Omaha, AG pa 0.20 sae 0.20 Sides B. H. Smith, Boston, Mass.. seal (O27 are 0.25 oer A. Valin, Ottawa, Cate 0.19 ae 0.19 iis H. J. W ichmaan, Denver, Colo......... 0.22 0.20 0.22 0.20 H. E. Woodward, Philadelphia, Pan ae 0.21 nas 0.21 Bee B. B. Wright, New orks NG Wee~ ser 0.20 0.18 0.20 0.19 150 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 The agreements and accuracy of these results, for a colorimetric proc- ess, are truly remarkable, and the method warrants further very careful study, with a view to determining its utility for extracts prepared by vari- ous methods, and in the presence of various possible interfering substances, for the purpose of testing its desirability as a method to be applied in special cases where the complete official method is not desired, or for confirmatory purposes. PEPPERMINT EXTRACT. Last year’s referee recommended that methods for the examination of ginger and other flavoring extracts be made‘the subject of future work. It appeared that a method for peppermint extract was especially needed, since the provisional method has been recommended for further study. The available proposed methods are all modifications of the precipita- tion method, originally devised by Mitchell for lemon extract. HORTVET’S MODIFICATIONS. In the modifications by Julius Hortvet brine is used for rendering the oil less soluble in the dilute alcoholic liquid. The method, as originally proposed, in the Journal of Industrial and Engineering Chemistry, Vol. 1, 1909, of floating the oil directly on brine, was carefully tried, and found to yield much too low results with extracts containing a low percentage of oil, but too high results for extracts containing a high percentage of oil. The reason for the low results on low grade extracts is the solubility of the oil in the alcohol-brine mixture. The explanation of the high results in high grade extracts is that the salt reduces the solubility of the alcohol in the water, so that part of it is taken up by the oil. In other words, the alcohol is partly salted out. Hortvet later modified the method further! with a view of overcom- ing this salting out effect by drawing off the alcohol-brine solution, either by means of a finely-drawn-out tube, or from the bottom of a specially constructed bottle. But this improvement was not suggested until after the samples and methods had been submitted to the collaborators. It was sent subsequently to a few men who were thought to be especially interested in extracts, with the request that it be tried on the peppermint extracts submitted. While only three responses were received, these were from men who have had much experience with extracts. 1 Personal communication. 1915] PAUL: FLAVORING EXTRACTS 151 HOWARD’S MODIFICATIONS. The modifications by C. D. Howard (Bur. Chem. Bul. 137, p. 76, and J. Ind. Eng. Chem., 1911, 3: 252) depend on the removal of the oil from the diluted extracts by means of a volatile solvent, which latter is removed before centrifuging. The original method, though provisional, was rec- ommended for study, two years ago, and it, and the further modifica- tion suggested by its author were tried by the referee, with quite satis- factory results. It appeared that some of the previously reported dis- cordant results may be due largely to inexperience on the part of the operators. It was thought that two procedures might be advantageously combined and very slightly altered, so as to make an accurate, safe, and quick method. The modification consists practically of the provisional method, carbon- disulphid being, however, substituted for chloroform and ether. The reason for substituting this solvent is that while its boiling point is higher than that of ether, it is much lower than that of chloroform. The details for removing the solvent are those of Howard’s later modification. CHITTICK’S MODIFICATION. The modification of G. H. Chittick contemplates collecting the oil, before centrifuging, in a measured quantity of nonvolatile solvent, and inferring the percentage of peppermint oil from the increase in volume or from the refractive index of the oil layer. This method was first read in the convention of American Dairy, Food, and Drug Officials during June of 1913, some time after samples had been sent to the collaborators. Its author very kindly supplied a copy in manuscript, but it was not practicable to obtain collaborative results on the method. It would hardly seem, though, that the relatively large volume of oil could be read, after the manipulation, with sufficient accuracy to yield very close results nor that the refractive indices would give more than approximate results. The results obtained by the various collaborators on extracts prepared in the laboratory from 95 per cent alcohol and finest selection of Michigan peppermint oil, are here tabulated: 152 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 Collaborator’s results on peppermint extracts. EXTRACT CONTAINING | EXTRACT conTarnina | EXTRACT CON- 0.5 PER CENT OF OIL 3 PER CENT OF OIL TAINING 5 PER CENT OF OIL sz |g we ‘sra))| = a 3 oh Sell detect] SNe, ilazexeaun woes. eS) ell herman are ANALYS) = : & A 28 ‘ 5 ‘ 5 T 3 g Dsl z 3 g os = I = aera rahe aeerats. ‘s] _m ng io) mn mag ag a ze | 22 | 82] 22 | £2 | 22 | 88 | #8 Ba | Be | 68 | a | 82 | be | BE | BS paged gies en | eg af ek fel 3 per cent| per cent} per cent| per cent| per cent| per cent| per cent| per cent 3.4 E. H. Berry, Chicago, Ill........ 0.5/Trace| Trace) 3.0 Fe 2.8 5.6 5.0 , 3.2 C. 8. Brinton, Philadelphia, Pa. {92)} ....J...... er eee Mie eed... H. C. Fuller, Washington, D.C..| 0.5)......]...... SiO) aes: ci: -rsccysre lee to | eee A.W. Hanson, Kansas City,Mo] 0.4)......J...... ECE epee erated eset alaccr oc H. A. Halverson, St.Paul, Minn.| 0.5 1 ; a} 0.2 3.0 3.0 3.0 5.8 5.0 N. Hendrickson, Omaha, Nebr..} 0.5)......]...... SHO eee aly ete el ov ees P. A. Holtzendorf, Memphis, IRENE ee aren neers ORAS ees ops] sere 320i ..ccrrellers svete ee eoee here C.F. Jablonski, New York,N.Y.| 0.6)......]...... Ce ARR Iasnieias etiog ed [scn.5 5 H. L. Lourie, New York, N. Y.... jo] aeallo Mane eallee NR eets Geteina coco 3 W. B.D. Penniman, Baltimore, 0.6 } Bini 1 Ko La eae Cr ey etl eo een ta A \ ORG ears aa pe 0 1 (ieee) eG PED S S000 S. H. Ross, Omaha, Nebr...... 106 inlanee ee } pave elrieleraiel bt aevavela | eres F. L. Shannon, Lansing, Mich..| 0.6)......]...... Boep | a ee ESAS 559 ci B. H. Smith, Boston, Mass..... OES elacnlipeees 2-Gl.5 5.55 else's oot acreel aaa A. Valin, Ottawa, Can......... gE Alls srakeesee a \ Seed beeen ss] sarees See H. J. Wichmann, Denver, Colo.| 0.4) none}...... 133 a0 } : { ; } H. E. Woodward, Philadelphia, BE earesaiesS emctendawien Meieneie OG Ras Scot [bs ciegits SHB) apsiete dls sarees] a, haha le eles a 0.6 3.0 B. B. Wright, New York, N.Y.. {ae Sey Sh sree ee... The following comments were made by the collaborators: E. H. Berry: The new modification is more satisfactory than any of the old ones. Hortvet’s new method gives fair results on samples containing 1 per cent or over of oil. An ordinary Babcock bottle and a drawn out tube are not satisfactory, a little oil is bound to stick and be lost. The special bottle may overcome this, but special apparatus is objectionable. C.S. Brinton: The results may have errors due to difficulty in reading the menis- cus. The method as prepared seems more satisfactory than that submitted last year. H.C. Fuller: I can not see any advantage in this modification over the procedure of adding salt and hydrochloric acid and reading the volume of oil separated. A. W. Hanson: A number of determinations were tried and we were pleased in obtaining results that agreed very closely. 1915] PAUL: FLAVORING EXTRACTS 153 Julius Hortvet: I recognize the fact that my new modification, even when carried out with a special bottle, gives rather low results on extracts containing appreciably less than 1 per cent of the oil. Our experience, however, shows that good results are obtainable on average commercial extracts containing oil in amounts approxi- mating to the standard requirements. H. L. Lourie: These readings are to the bottom of the meniscus. Reading to the extreme top gave 0.8 and 3.4 respectively. This method does not impress me as capable of giving accurate results. If the distillation of carbon bisulphid is con- tinued too long there is loss of peppermint, if too short a time there is contamination with bisulphid. W. B.D. Penniman: Both Howard’s method and this modification give the same results and require about the same time. Howard’s modification has always given satisfactory results in this laboratory. A. Valin: I experienced a little difficulty in telling when the carbon disulphid was wholly removed. H. J. Wichmann: I have tried Hortvet’s method and have substituted solutions of various other salts, but without success. By the use of a drawn out tube there is a possible loss through oil sticking to the tube on its withdrawal. Lourie’s objections must necessarily apply to any method depending on the removal, from the oil, of a volatile solvent. As a matter of fact, however, peppermint oil has a relatively high boiling point, and the danger of loss by volatilization is easily overestimated. Certain it is that there is ample leeway between the two dangers mentioned, to enable any opera- tor to obtain good results with very little practice. This is shown by the results reported by the collaborators. Hortvet’s original method, which has the advantage of ease of appli- cation and of time consumed in carrying out the operation, is undoubtedly very gootl for extracts approaching standard strength, and will be valu- able for a preliminary sorting out of a series of unknown extracts, or in cases where it is merely necessary to decide whether the products are of standard strength or not. For the determination of the actual oil con- tent in low grade extracts, the former method is better suited, and in all cases yields results remarkably close to the truth. RECOMMENDATIONS. It is reeommended— (1) That the results reported herewith on known samples of vanilla extracts be submitted to the Committee on Standards, for consideration in connection with such other results as are available in establishing standards. (2) That Folin’s colorimetric vanillin method, as described in the Journal of Industrial and Engineering Chemistry, 1912, Volume 2, page 670, be further studied on extracts prepared in various manners, and on other products containing vanillin, especially in the presence of substances which would interfere in the provisional procedure, to test its value as a 154 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 method for use in special cases where the gravimetric process is not ap- plicable, and for confirmatory purposes. (3) That Wichmann’s qualitative method be given further considera- tion in connection with Folin’s method, and as a preliminary test. (4) That the following slight modifications of Howard’s method for the determination of oil in peppermint extract be adopted as a provisional method: Pipette 10 ce. of the extract into a Babcock milk bottle, add 1 ce. of carbon disul- phid, mix thoroughly, then add 25 ce. of cold water and 1 ce. of concentrated hydro- chloric acid. Close the mouth of the bottle with the thumb and shake vigorously for not lessthan 1 minute. Whirl the bottle in a centrifuge for 6 minutes and remove all but 3 or 4 ce. of the supernatant liquid by means of a glass tube of small bore and aspiration. Connect the stem of the bottle with a filter pump, immerse the bottle in nearly boiling water, start the pump and shake gently. When the carbon dis- sulphid is practically all removed, the oil will float entirely on the surface of the watery layer. Then shake violently and immerse in boiling water for a few seconds. On disconnecting from the filter pump no odor of carbon disulphid should be detected. Cool, fill the bottle to the neck with saturated salt solution at room temperature, centrifuge for 3 minutes and read the volume of the separated oil from the top of the meniscus. Multiply the reading by 2 to obtain the per cent of oil by volume. THE DIRECT DETERMINATION OF VOLATILE OIL OF CLOVES BY DISTILLATION WITH STEAM. By Junius Hortver. The difficulties attending the accurate determination of essential oil in cloves and other spices have been the subject of frequent discussions dur- ing recent years. Our present official method for spices is essentially a procedure for the determination of total and volatile ether extract, and no analyst regards the so-called volatile constituent as an accurate result for essential oil. These remarks are especially applicable to the analysis of cloves. McGill,! Collins,? and Brooks* have suggested modifications of the procedure and also new methods, but none of these seem to have ad- vanced to a condition suitable for general application. The high tem- perature to which the total ether extract is subjected in carrying out our present method of analysis is one of the conditions which has been much criticised. It is conceivable that a temperature running as high as 110°C. may cause a loss of total ether extract considerably in excess of actual volatile oil present; the result obtained may include moisture and decom- posed organic substance of unknown composition. 1 Inland Revenue Dept., Ottawa, Can., Bul. 252. 2 Analysis of Ground Cloves, Swarthmore, Pa., 1910. >The Spice Mill, Sept., 1911. 1915] HORTVET: DETERMINATION OF VOLATILE OIL OF CLOVES 155 The determination of essential oil of cloves by direct distillation with steam has been proposed, and such a method has been brought forward by Girard and Dupré. The method consists essentially in mixing a weighed amount of spice with water, subjecting to distillation, and receiving the distillate in a graduated cylinder. Itis stated that the volume occupied by the essential oil which is immiscible with water and separates out can be read off and its contents roughly determined. For a more accurate determination the mixture of oil and water may be extracted with petro- leum ether, the ether evaporated, the residue dried at room temperature, and weighed. R. Reich! has described a procedure based on this latter method, in which he distills 10 to 20 grams of the spice in a specially-constructed ap- paratus which provides for the passage of the steam through the loosely- packed sample placed in a capsule intermediate between the steam gen- erator and the condenser. The distillation is continued until 600 to 800 ce. of distillate have been collected. The distillate is transferred to a separatory funnel, saturated with common salt, and extracted with sul- phurie ether or petroleum ether, the latter solvent being preferred. The time required for the distillation is said to be 14 to 2 hours, and the entire determination, including distillation and extraction, may possibly be com- pleted in 3 hours’ time. The principle of the method involves certain features which are attractive and which indicate a possibility of an accurate procedure for the direct determination of volatile oil m cloves. Accordingly, the method has been subjected to study, and experience has developed certain modifications which seem to constitute improve- ments chiefly in respect to simplicity of detail and time of carrying out a determination. DETERMINATION OF VOLATILE OIL OF CLOVES. Apparatus. The apparatus used is the one described for the determination of volatile acids in wine and other liquors (J. Ind. Eng. Chem., 1909, 1:31). It consists of a 350 ce. spherical flat-bottomed flask provided with an elongated wide neck, into which is fitted a 75 ec. cylindrical-shaped flask provided with a siphon-like side tube. Into the latter flask is fitted a small funnel with a stopcock and a de- livery tube with safety bulb leading to a condenser. Procedure. Place 2 grams of the spice in the inner tube of the apparatus and add 25 ec. of a 20 per cent salt solution; place 250 cc. of a 20 per cent salt solution in the outer flask, close the apparatus, and attach to the condenser. Apply heat, leaving the side outlet tube of the outer flask open until boiling has fairly begun. Then close the outlet tube and force the current of steam through the mixture of cloves and brine in the inner tube. Continue the distillation until 225 cc. of distillate have been collected, wash the distillate into a separatory funnel with 35 cc. of sulphuric ether, shake the mixture well and allow time for a good separation. The separated ether will have a somewhat cloudy and emulsified appearance. Make 3 more ex- 1Zts. Nahr. Genussm., 1909, 18: 401. 156 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 tractions of the distillate, using in succession 25, 15, and 10 ec. of ether. Wash the combined ether extracts with an equal quantity of water, thereby removing the cloudy appearance of the ether; a second washing is sometimes advisable. Pass the washed ether through a dry filter and wash with a little ether into a previously weighed dish, place the dish and contents in an air oven at room temperature and allow the ether to evaporate spontaneously. When the last trace of ether has disappeared weigh the dish and contents. Results obtained by this method, compared with results obtained by the present official method, are shown in the following tabulation: Results on determination of volatile oil of cloves. a TOTAL VOLATILE FIXED WEIGHT OF | - oy arie aos LABORATORY NO. psa aee Saran Shae paetetn OIL ENCES per cent per cent per cent grams per cent per cent AS cee te aera terete steers 27 .80 19.23 8.57 0.3595 | 17.97 —1.26 Ibs Pe oda sans one ab oo 25.81 18.41 7.40 0.3583 | 17.91 —0.50 ibe arrose monmomen cor 24.86 16.11 8.75 0.3129 | 15.64 —0.47 TST S WA Ee. Beit. esceee isl oor 26 .87 17.73 9.14 0.3497 | 17.49 —0.24 1800): .itdcsrcte seiciseeooee 27 .45 19.03 8.42 0.3642 | 18.21 —0.82 SOG oo etek. yetencto vet ore eat 23 .37 15.61 7.76 0.3162 | 15.81 +0.20 1 LS ayes ih ie ane eg ee 25.25 16.93 8.32 0.3339 | 16.69 —0.24 18035 2). see hie ace sese cies 24.05 16.02 8.03 0.2908 | 14.54 —1.48 hs ag to cree GOO Ean aS 26.61 Ufa ly/ 9.44 0.3032 | 15.16 —2.01 TS20 TOR aa Aye hes yarns 25.10 17.03 8.07 0.3065 | 15.32 —1.71 1 S2Te seen erate meee 28 .86 20.72 8.14 0.3752 | 18.76 —1.96 S30 i foerotcbeles seaushorseteeieteis 25.75 17.80 7.95 0.3283 | 16.41 —1.39 RSI See ane ee Ne 28.48 20.73 7.75 0.3590 | 17.95 —2.78 UPS eros DOAcaoD OO 26 .27 17.05 9.22 0.3168 | 15.84 —1.21 NSS ioe cpietempers cise wane 24.76 16.72 8.04 0.3168 | 15.84 —0.88 SAO Macca lst ststecictraes 28 .54 19.93 8.61 0.3736 | 18.68 —1.25 OS eee tee rosy eee 27 .80 19.15 8.65 0.3622 | 18.11 —1.04 B28 hs ysttvcterocsor nis, neuter 25.35 16.98 8.37 0.3184 | 15.92 —1.06 SSG cers lee ne le coke 25.84 16.52 9.32 0.3038 | 15.19 —1.33 Maximum.............| 28.86 20.73 9.44 0.3752 | 18.76 —2.78 Mami ums ee os os cae 23 .37 15.61 7.40 0.2908 | 14.54 —0.24 The distillation method carried out as described gives results below those obtained by the official method, the differences varying from 0.24 to 2.78 per cent. Some of the high discrepancies are accounted for by the fact that the volatile constituent of the total ether extract was accidently subjected to a heat considerably above 110°C. At any rate, it is not claimed that the results by the official method afford criteria for judging results by the method of distillation. The best check on the distillation method would consist in other determinations made by the procedure modified in various ways. Five repetitions of the method on a given sample gave results varying from 17.90 to 18.06 per cent. By continuing the distillation so as to collect distillates beyond 225 ce., measuring 100 ec., 50 ce., and 50 ce. respectively, a total gain in volatile oil was obtained 1915] HORTVET: DETERMINATION OF VOLATILE OIL OF CLOVES 157 amounting to only 0.02 per cent. Further attempts to obtain a greater yield of volatile oil by further extraction of the distillate with ether were found to give no increase in the results. The spice residue remaining in the inner tube after each determination was found to have no odor resembling that of cloves. The saturation of the distillate with salt as recommended by Reich has not been found neces- sary or advisable. The addition of a small amount of salt solution to the sample in the distilling tube has been found to improve the condition of distillation, maintaining the volume of material uniform throughout the entire process. An improvement in the apparatus has been effected by means of a perforated enlargement at the lower end of the inner steam delivery tube. The four or five perforations thus provided serve to break up and scatter the steam jet, thus effecting a more complete mixing and contact of steam with the spice. The rapid evaporation of the ether from the volatile oil has been found inadvisable, whether by contact with live steam or other artificial heat, a very decided loss of oil resulting from any procedure of this kind. Under various conditions this loss has been found to amount to from 0.3 to nearly 2 per cent. Furthermore, it has not been found advisable to dry the residue over sulphuric acid, the loss overnight being especially appreciable owing to the fact that the oil is taken up by the acid. Placing the residue after complete evaporation of the ether in a calcium chlorid desiccator has been found to effect a complete removal of mois- ture, so far as can be judged by the appearance and other properties of the residue in the flask. A number of details of this determination have yet to be worked out, but enough experience leads us to the belief that a determination of volatile oil of cloves by a procedure like the one which has been described is a proper method for such a determination, inasmuch as it is believed that the actual oil can be recovered in a sufficiently pure condition for weighing and moreover represents the actual volatile oil in the spice. If this method as applied to cloves proves to be satisfactory, the same procedure may be extended to other spices. A motion made by V. K. Chesnut that two additional associate referees be appointed on medicinal plants and drugs was carried. A resolution by L. L. Van Slyke for the appointment of a committee of three for the study of vegetable proteins was adopted. 158 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 PRESIDENT’S ADDRESS. By G. 8S. Fraps. It is the object of this address to present briefly the important recent advances made in agricultural chemistry. In so doing, it is not my in- tention to go back one hundred years, or fifty years, or even to the period included in the memory of the veterans of this association; but to consider only such a period as is within the memory and the experience of a young- ster like myself. Agricultural chemistry is so closely interwoven with the other sciences which have been applied to agriculture, that it is practically impossible to disentangle them. Hence to a certain extent the progress of the chem- istry of agriculture is closely related to the progress of other agricultural sciences and to agricultural science, in general. The contributions of the chemist to agricultural science have been so many, so varied, and so im- portant, that for a long time the sciences applied to agriculture have been termed “agricultural chemistry.” This period is passing, and the term “agricultural chemistry” is being more restricted in its significance, but the field is still broad, and the harvest bountiful to the worker who seeks to garner the grain of knowledge. There has been a tendency in some colleges to discontinue the teaching of agricultural chemistry, and to divide the subject matter between the agronomist and the animal husbandman. It is a serious question whether such tendency is in accord with the known laws of specilization in science. There is no doubt but that, as time goes on, the agricultural chemist must specialize more and more in one of these fields of work, but there is a difference between the specialization of the scientist in his own field, and the attempt of other branches of agricultural science to take over the work of the chemist, or the chemist to take over other branches of agricultural science. As I see it, both the agronomist and the animal husbandman have their special problems. They must have their special training in their own fields, and while this training must include some chemistry, it is not sufficient in quantity to make them into chemists. On the other hand, the chemist must be first of all a chemist. The agricultural chemist must have knowledge of soils and animal nutrition, but he should have predominant chemical training and chemical methods of thought. The agronomist and the animal husbandman undoubtedly need the aid of the chemist in the solution of their problems, but they should not seek, at one and the same time, to be both agronomist and chemist. The result of such an effort is either an agronomical chemist, or a chemical agronomist. It often results in the chemist becoming also the agronomist. What agricultural science needs is the highly-trained agronomist, working, where needs be, in codperation with a highly-trained 1915] FRAPS: PRESIDENT’S ADDRESS 159 chemist who has perhaps specialized in soils and fertilizer chemistry, each assisting and aiding the other. The same is true of the animal husbandman. We need the animal husbandman, highly-trained in his field and with a full knowledge of its peculiar problems, working in codperation with the agricultural chemist, highly specialized in the chemistry of animal nutrition. In this way, we shall avoid those errors which we so often see when a man enters into a field outside of his special trainimg—errors which the specialist immediately recognizes. The truth of the matter is, that the chemist has made such great contributions to the field of agricultural science, that the agronomist and the animal husbandman have, in many cases, not been able to see their own peculiar problems, but have emphasized the chemical side of the subject. They have not wholly found themselves. In some institutions, agricultural chemistry is no longer taught. This, we believe, is a mistake. The student needs a thorough grounding in the entire field, such as is given by the agricultural chemist, and he needs to look at agriculture, for a time, from the point of view of the chemist. Specialization should come later. These matters will adjust themselves in time. We need not fear that the science of agriculture will ever be without the need of the agricultural chemist. Our ranks have not thinned, but each step of progress has rather added to our numbers. The Adams Act of March 16, 1906, for example, which is one of the most important events in the recent history of agricultural science, has increased the number of agricultural chemists, as well as the number of other agricultural investigators. This act is important, not only because it increased the number of scientific agricultural workers in the experiment stations and their facili- ties for investigation, but because it affords to the experiment stations opportunity for fundamental research work. The passage of the Adams Act indeed marked an epoch in the history of agricultural science. The experiment stations had previously done much valuable work, and accu- mulated much data, a fact which the passage of the Adams Act itself recognizes, but they had such large demands upon them for immediate and practical information, that they had little time for the investigation of fundamental things, which are no less practical in their final application, but require more time and patience, and are less obvious in their practical applications. Under this act, the experiment stations not only may, but must, conduct research. Fundamental and continuous work may be done upon projects which have no present popular appeal, though no one can predict the ultimate effect of such work. The result of the Adams Act has been an increase in personnel and in facilities, and has aided in creat- ing a demand for more highly-trained, research assistants. It has also tended to raise the standard of scientific publications of the stations. Thus, with the passage of the Adams Act, the experiment stations entered 160 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 upon a new period of their existence, one in which fundamental research becomes a much greater part of their work than has been the case in the past. It is true that some directors of stations, and some governing boards, do not yet understand the true significance of research, or the qualifica- tions necessary to pursue it. It is true that some station men do not, in their publications, give proper reference to previous work, which may have anticipated their own. It is true that in bulletins and reports of directors, we sometimes find claims of credit for work which are exaggerated or perhaps the credit belongs elsewhere, claims which are hardly pardon- able, even after making all possible allowance for naturally exaggerated opinions of one’s own work. Such things will pass away. We need more criticism of our agricultural publications—not destructive criticism, but friendly criticism, and friendly controversies over disputed points. Criti- cism of the proper kind is a stimulant to good work, and aids in pruning away excrescences such as those mentioned. The Adams Act created a demand for men capable of research in agri- cultural chemistry, and other lines of agricultural science. Research is not an ordinary qualification, even in young men just graduated from col- lege. The ability to do research work must be founded upon a natural ability and inclination toward such work, developed by broad general training and wide knowledge of some particular science, and by an ap- prenticeship under one who is himself a master of research. This ap- prenticeship may be during a course of work and study for the degree of Doctor of Philosophy; but it may also be in the process of regular station work under some eminent station investigator. We must recognize the fact that all men capable of research have not been able to secure the doctor’s degree, even though they have done equivalent work. The ability to do research work may be developed by study and training, but it can not be created. The Adams Act thus marks an important step in the progress of agri- cultural chemistry, other agricultural sciences, and agriculture as a whole. Perhaps equally as significant was the passage of the National Food and Drugs Act, approved June 30, 1906. Taken in a broad way, the passage of this act was one of a series of events in the reaction of the people against dishonest commercial practices. It has become evident that the people will no longer tolerate practices which have crept into use, which are morally wrong, but were formerly considered as all right because they were business; practices which deceive the buyer or give unfair advantages in business competition. Business has been a species of warfare, but Just as it is now contrary to the laws of civilized warfare to kill women and children, and burn private dwellings, so it is becoming contrary to the laws of business warfare to cheat women and children, and to deceive the 1915] FRAPS: PRESIDENT’S ADDRESS 161 purchaser as far as possible. How much the agitation for the pure food and drug law had to do with this moral awakening, no one can say, but no doubt this crusade of twenty-two years had much to do with it—a crusade by an agricultural chemist, Dr. Harvey W. Wiley, for many years chief of the Bureau of Chemistry, secretary of the Association of Official Chemists from its organization until only a little more than a year ago, now our honorary president—for whom all of us have a warm place in our hearts. The Food and Drugs Act has resulted in a material clearing of the at- mosphere with respect to the naming, labeling, and adulteration of foods, drugs, and feeds. We now have very clearly defined the objects of such a law. These are, first, to prevent the sale of any unwholesome or dele- terious substance, and second, to ensure that the goods delivered to the purchaser shall be exactly as represented. These principles have been made clear, not only with respect to foods and drugs, but also with re- spect to feeds, and feed manufacturers are beginning to realize that a mixture of bran and screenings may no longer be sold as bran, or a mix- ture of corn bran and corn chops, sold as corn chops. There are some feed manufacturers who have not yet read aright the signs of the times, as, for example, some of the manufacturers of cottonseed meal, who contend for the authority to sell a mixture of meal and hulls under the name of cottonseed meal, but undoubtedly the time will come when this matter will be made clear. This association has played an important part with respect to food adulteration. Before 1900, there was one referee and one associate on this subject. At the 1900 meeting, provision was made for 14 associate referees, and there are now 21 associate referees. In addition, we have our Committee on Food Standards, which has done valuable work. In the matter of cattle feeds, their analysis and adulteration, it appears this association has done little in recent years. The analysis and control of these feeds are yearly assuming a greater importance. There should be a referee and an associate referee on the adulteration of feeds and the detection of adulterants. We have no official methods on this phase of the subject, beyond the ordinary analysis. The method for crude fiber should be thoroughly studied, and perhaps modified. The clause which permits filtration through cloth should be eliminated. The estimation of crude fiber is becoming more and more important, for by its use we can de- tect more easily the addition of materials rich in crude fiber, to concen- trated feeds. The estimation of crude fiber, for example, shows much more clearly the probable quantity of cottonseed hulls in cottonseed meal, or of rice hulls in rice bran, than does any estimation of protein and fat. Striking progress has been made in recent years in the survey and mapping of soils. In this work, the Bureau of Soils is easily the leader. 162 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 1 There is a tendency in some quarters to regard the survey, mapping, and analysis of soils as an end in itself. It is true that such work is highly important, but it should also be regarded as a basis on which to make further soil investigation, so that we may become fully familiar with the properties and characteristics of each type. In a sense, the soil survey should be regarded as the beginning of soil studies. In other respects, our knowledge of soils has been increased by recent investigations. We now know more concerning the nature and constitu- ents of the organic matter of the soil, and something more concerning its biological properties. We also know that, on an average, the needs of the soil for fertilizer nitrogen in pot experiments is related to the total nitrogen of the soil. We know that the active potash of the soil is related to the average needs of the soil for potash in pot experiments, and that plants have the power to exhaust the active potash, and to take up more potash than they need. We know that, on an average, the active phos- phoric acid of the soil is related to the needs of the soil for phosphoric acid in pot experiments. The relation of the pot experiments, and the analysis, to field needs, must be worked out. Soils also deviate from the average, as regards their plant food content and behavior to pot experiments; such deviations must be studied, and their causes ascertained. There is much to be done, but progress is being made. In the field of animal chemistry, decided progress has been made in re- cent years. We must now recognize the possibility that, in digestion, proteins of different kinds may be split into different products, some of which may be unfit for use as structural material in building up animal proteins, and so must be discarded. We know that this is possible, but we have not yet secured positive evidence that such occurs with any of the various proteins fed to domestic animals. Such studies may be expected in the future. It has been shown, without doubt, that the digested material of different feeds have different values to the animals. One pound of digestible ni- trogen-free extract in corn, has a much greater value than one pound of digestible nitrogen-free extract in straw. The fact that there is a differ- ence in the value of the digested nutrients of the same class but from different feeds has been clearly shown by the work of Kellner and of Armsby. There is no doubt about it. It is a step forward to recognize the differences in the values of the digested nutrients, and to adjust our tables, our rations, and our calculations accordingly. There is abundant room for work along this line, but enough work has already been done to justify this advance. Nearly every American book which deals with the feeding of animals still assumes that the digestible nutrients of one feed are equal in nutritive value, pound for pound, to the digestible nutrients of the same class in any other feed. These books must be rewritten, and 1915] FRAPS: PRESIDENT’S ADDRESS 163 adjusted to our latest advances in knowledge. This advance will, to a certain extent, reconcile these discrepancies between the effects of feeds or of rations in feeding experiments which, under the old standards, should have apparently the same nutritive values. We are now able to state the nutritive value of a feed in terms of three factors; its bulk, which satisfies the hunger of the animal; its proteins, which repair flesh or tissue, or which, in excess, may be used for fat or energy; its fat-producing value, which is its ability to furnish the animal with heat or energy or to form fat. The fat-producing value of a feed or nutrient is determined experimentally. First, the fattening animal is fed aration which produces a slight gain of fat, and the gain of fat is measured by determining the income and outgo of carbon and nitrogen. Next, the nutrient of feed is added to this ration. and the gain in fat again deter- mined. The difference in the quantity of fat produced is due to the added feed or nutrient. The results of such work may be readily compared with calculations based on the assumed equality of the same group of nutrients in different feeds. While the calculated value of peanut meal or linseed meal is prac- tically equal to that found, the value for a wheat straw is only 20 per cent of that calculated, of meadow hay 54 per cent, of rye bran 79 per cent of that calculated. It should be clear that the recent advances in the chemistry of animal nutrition, compel us to modify materially tables of feeding values, rations, and methods of calculations. There is opportunity for usefuland valuable work along the lines of determining exactly the productive values of feeds and nutrients, and such work may be expected in the future. In the thirteen years of the twentieth century, the progress of agri- cultural chemistry has been such as to satisfy even the pessimist that we are moving forward. Our facilities for scientific investigation have been increased by the Adams Act. Our supervision over foods, drugs, and feeds has been enlarged and rendered more effective through the Federal Food and Drugs Act. We have made great progress in the survey and mapping of soils, and in our knowledge of their properties and chemical composition. The science of animal nutrition has made such advances as to render it necessary to revise almost all books dealing with the sub- ject and to modify our methods of stating the nutritive value of feeds and our methods of calculating rations for feeding animals. These have been the four chief lines of advance of agricultural chemistry in recent years. The members of the Association of Official Agricultural Chemists may well take pride in the part they have taken in the progress that has been made. : The Association adjourned until 2 p.m. INDEX TO VOLUME I, NUMBER 1 Adams reportonidistilledispirits sepa eee ee Eee eee one eer. 143 Ammonium carbonate, effect on determination of humus, paper by McIntire EieX0 lil b [hich aoa geet atten aninosd eRe oie Wo. noo ne pao aee 44 citrate solution, neutral, preparation, paper by Patten and Marti, RELCECINC EN: coees tials a sl iat e aN ATES ee Terao Ne Eye 17 Andrews, paper on fruit jellies, reference..................0. 0. ccc cee eee cues 130 paperonstruitsjuices\ireference:j...cata-s aa eee eer 130 Arsenate, lead, water-soluble arsenic content, report by Averitt............... 74 Arsenic, water-soluble, in lead arsenate, report by Averitt.................... 74 Auditing, committee, appointment and personnel.........................-05. 59 Averitt mpaperionulime-sulphur solutions’ ...-cs2 cc eee eee eit eens 95 Neportmonwminsecbicldese. cess seal sees se ee cae eee 59 Basic slags, phosphoric acid content, report by committee (Williams)......... 102 report by Patten and Walker............ 8 BEST PLE POLLO Vas Ewil eye see ais oysters cvartiey cle alerts arrest eeie eto ale eter dGactere ae ences create 138 Blair and McLean, paper on lime requirement of soils....................2.0- 39 Bosworth, paper on sodium citrate for determination of reverted phosphoric EYGHGL SAD KEIRENAECILE, a clde dk OE EEO ERG One on OMtericeR Belaamerd cuanto Ges Gonnne 17 Carbonate, ammonium, effect on determination of humus, paper by McIntire Aral IBLE hier dd ton Dene Cem Gn eo Cee EEE ERNn ead ca ed ade oters cote 44 Citrate, sodium, for determination of reverted phosphoric acid, paper by IBOS wont eROLELOM CO epee al tens saree n erase 3 ocie oae REESE Stor eiepe eters 17 solution, ammonium, neutral, preparation, paper by Patten and Marti, MEL ELEN COM yar har eer foe oS SST Leyes NSD Seep ERO en ree IS 17 @loves*; oll spaper Dy PELONEVEL 5. ejc'2.acers.9:6/ 5,410) syerheidvarsisinlsieje = aa lopeuatevero ms sus eccke mesiefelare 154 Colors; recommendation by Mathewson ..... 2.2. -c ons secon noeceec eee ser 120 RSTO Loh IM IeMi AT Gao) lea ae ORE ae Roaoee a ae ees Ret oe an aocddoads sean 113 Committee A on recommendations of referees, report..................0.202005 100 Committcesyappominentsand) personnel: s.c--40- cas eee eee nae eee 59 Curry and McIntire, report on inorganic plant constituents................... 55 Dis tilledtspintismepontiby PA sIMNSs.o.21.. 2)-'hescmiecie terse ictoeier sferersicle le electors aisle 143 Drugs and medicinal plants, appointment of two new associate referees....... 157 Extracts, favoring, recommendations by Paul........................+eseeeeee 153 MEPOLUEID Va era les so vehees rh dlovorctore tava overs ous payalene Bidar aves SLA SeNTO IVs ots ereetole 146 Feeds and feeding stuffs, adulteration, appointment of new associate referee 107 Feldspathie fertilizer, potash content, paper by Miller and Vanatta........... 26 Flavoring extracts, recommendations by Paul..................-..-0eeeee neces 153 MEDOT bi YP AULS <.croes Seco. esteae eee io tecctenee eke eee 146 Hood adultentiony report bys Elortvetean dere aeiscnin ciel scia. 1 eriatiesieiaerctats 110 standards, report by committee) (Frear). 0.2... saseacesneece + lsnoe ee 108 1€5 166 INDEX IME HO ovo lou Hist tK6 ls ht enemOR BEM aac son ooe oo Oa poodeOn oS Gun Moacereaneameys 158 TEPOLEVOD SOUS. fsa eis. F evsrsr siege epats tere enagers stele re ecorcncdotche eecenereIe ete eel rere a 33 Frear, report by committee on food standards............................0--. 108 Bruit jellesspaperibyy Andrews, TeleLenCe sj .crit terials ere aie 130 hurces;papernby, Andrews, TLelerencen-en ee ete sie ee a eee eee eee 130 products, recommendations by. Gores eee eee eee lee eel eee 130 report by Gorevacnd ase eauwonceye smtars easiner tera re CTT 120 Goodnow, Teport Onc VANE arses er see see eels ieee eile eee oie 145 Gores reportion iruitiproducts:)2 ee eeeer ee eee ace a eee 12 Hardy and McelIntire, paper on effect of ammonium carbonate on determina- tion: Of HUMMUS wis hccneee eer sae ein ae Seo Cee ee eee 44 Hare, report.oninitrog ens cnsserte sei tet psy uses wick oes eee chee See eee 17 Hartmann) epont, Oni win see de ere et etre al eer taney canes eaten ci wie ike. sisvtgees 131 Haywood, report by committee on editing methods of analysis..... steata tiers Meet 108 Hortvet, paper on oil of cloves............... Piet Porte are neds Gece eee 154 reporton Loodyadull erations: qe. eee ee eee eel alee ciara ese 110 Humus, determination, effect of ammonium carbonate, paper by MeIntire and aE Eds heennea Seam eid ror Dau SOA oe 44 DAaper:D yi Smith yee ee Lek ct ee nee Earn ees ee 46 LEPORM bY SL TAPS sehen ek cole os become cei dor eeeeelonee emcee 35 Insecticides, recommendations by Averitt. ./.05.0.-52 5552-55. ..-- 120.2 e- = see 75 recommendations by, CommitteetAV=y..-.-. sacs es sou. cece 101 TEports Dy AVerney npc cya sre Warne ste ape ro Sorbonne nice @ cro eerie Means 59 Jarrell paper on determination! of potashe...20 4247-92 as ee oe eee 29 Jones) paper on: lime requinement/onsoulSss. saan oe eee cise settee eee 43 Lead-arsenate, water-soluble arsenic content, report by Averitt............... 74 ibhimey requirement) of soils\notelby Veitch ess --eesy- eee atari el ete eee 44 paper by, BlaintandsMicleans..s5- ee eeeeeeer aaeeee 39 paper iby JOnesty ss aesivacier asa ce ee tice honta ee eee 43 iinme=sulphur solutions, paper by; Averiute ane jeritct) eee sere eiicteterl ieee 95 paper by Roark cached ctnins eres enim col aeer eee 76 recommendations by Averitt.......................... 75 recommendations by Committee A.................... 101 TECOMMENAALIONS bys WOALK= seen eerie ae 94 report: by “Awvierttsio. ch his acct Sepateers ati ae eee 60 MeDennell, report on determination of potash....................0.-.-eeeeees 22 MelIntire and Curry, report on inorganic plant constituents................... 55 and Hardy, paper on effect of ammonium carbonate on determination OLTHUMUS Sere sic hod cro rersks aescnere WS Here lel etter ee OLE 44 McLean and Blair, paper on lime requirement of soils...................... 39 Marti and Patten, paper on method for preparing neutral ammonium citrate solution, ‘reference: a sscctetca ine sot eek Haan eee aoe Eee 17 Mathewson; treport om Colors cj. ccijae:sisreeissciesd aseisntrele icers eee stteie sie ies syjceee ss ae 113 Medicinal plants and drugs, appointment of two new associate referees........ 157 Members’at:1913convention:).ticeruacsehaeeio Gen Gaiden asia nec i er acre 1 INDEX 167 Methods of analysis, editing, report by committee (Haywood)................ 108 Miller and Vanatta, paper on potash in feldspathic fertilizer.................. 26 National Canners Association, invitation to smoker........................... 22 Nitrogen, Kjeldahl method, appointment of new associate referee............. 107 recommendations by, Committee Avan...) .ccee ese cess cases ose ee 100 Tecommendations DyPelares ja) cae OCT ae ete ikon 22 Keponbiby sare stata eters ots eerie ke eee OE Oe eee reece 17 Nitrogenous compounds in soils, recommendations by Plummer............... 54 Ke) CONC) lonvadedibbasveVere on Ga wee oe seep anoesnderue 49 Nominations, committee, appointment and personnel.......................... 59 Oil Ncloves;paper bypblontvetee ci s-r2 5 <1 nts rm sis, oe ete toe clean ioe 154 Patten and Marti, paper on method for preparing neutral ammonium citrate SOLMIIONPRELERENCOms a yee ce een oars ate ales ee 17 andiWalker reporiwonsphosphoricacidas..c:c «2-02 sac eee eee ee 8 Ral yreportion’ flavorinevextractsmees cess secis-iace- seco 6 sc oe siavctetee ass tele cialis 146 Phosphoric acid, in basic slags, report by committee (Williams)............... 102 report by Patten and Walker................. 8 recommendations by Committee A............. 100 recommendations by Patten and Walker ...... 16 report by Patten and Walker.................. 8 reverted, determination by use of sodium citrate, paper by Bosworth, reference..... 17 Plant constituents, inorganic, recommendations by Committee A............. 101 report by MecIntire'and Curry....:-....--:--.:+ 55 Plummer, report on nitrogenous compounds in soils.........................-- 49 Potash, availability, in feldspathic fertilizer, paper by Miller and Vanatta.... 26 report) by Wanatta-«--..-- 4... 24 determination; ;papenbyal ances hres aes sclanee anion erie 29 report pygvic@onnelleaee sae ans cjss.es'm eyes sco k plan ee 22 recommendations by McDonnell....................... 24 recommendations by Committee A..................... 101 residents: address; by. Hrapserepmemetseee conse cs on nace emcee etlaicreera dines 158 Publication of proceedings, announcement by secretary....................... 17 Resolutions, committee, appointment and personnel........................-. 59 Riley, report On DEOL. sence cieaerer te eisiole steele le as, oe citycier ale’ aa creas cline ate ajenemsyayale se 138 Riosrk, paper on lime-sulphursolutions-a- caste cee cicnts aed cee ania ciate 76 Ross, report by Committee A on recommendations of referees................. 100 Skinner, Teport) ON Water etcmerr eerie come csc aus erie ais aeisesrlerre a eannoe 97 Smith, paper on’ determinationiol humus: 2-0... ascccce ec cssce + sageseceas ce soe 46 Sodium citrate for determination of reverted phosphoric acid, paper by Bos- worthy releren cencretmeceer: siete nts ie raren rae ee ofote ei sasloaie cs tamakele ae niet 17 Soils;tacidity report by xbeaps ee sce she aed eee cele See eee Rees 33 alkali, appointment of new associate referee.................2.0...2000: 107 lime;requirement; Note byav eltch sas-penme +152 serine cls ele aratiersoietnee eee 44 paperby, blairsangdwNicheans=asseee as senna eee 39 paper by-J Ones? a. seeise saree) tnvoveeie sss ciceraisiiawe sels oe 43 168 INDEX Soils, nitrogenous content, recommendations by Plummer..................... 54 reportiby elummer..f-.c chic aoe e ee eee eee coe 49 recommendations, bys@ommitbee Avg. css on -e ee eee ae 101 recommendations by: PTapSi-.cn-eercenee ce cece cee ee eee 39 TEPOLUD YE TADS co oe Sinks games ee eee a ee eee eee eer cis 33 Standards, food, report by committee (Frear).....................0:--eeeeeee: 108 Wanatta report on availability of potash-pre.s---- ee ace eee eee eee eee 24 and Miller, paper on potash in feldspathic fertilizer.................. 26 Van Slyke, report by committee on study of vegetable proteins............... 109 Vegetable proteins, study, appointment of committee......................... 157 report by committee (Van Slyke) .................. 109 Veitch, note on) lime:requirement ofisoilssse5- ss. 0--5 2 ee seen ee 44 Vinegar, report by“ Goodnows.s a0. > see soca oe te Sere oo ele igemicles icine Serene 145 Visitors ab L9lSiconventiones-..jecies oo cee see ete eee erences faint eae Cee 1 Walker and Patten, report on phosphoric acid.........................--.2.- 8 Water, recommendations by Committee Aj: ::..-2.2-.020.cses ses eee rere ues 102 report by Skinner ?.).2. 092: decade yee hee satis SO eee OES Chee aie eee 97 Williams, report by committee on phosphoric acid in basic slags.............. 102 Wine; report, by Hartmann. soo. so eecd- see eas sce sts ie ew sane eterernieeioets 131 PROCEEDINGS OF THE THIRTIETH ANNUAL CON- VENTION OF THE ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS, 1918. SECOND DAY. TUESDAY—AFTERNOON SESSION. W. A. Withers, chairman of the committee appointed to convey an invitation to the Secretary and Assistant Secretary of Agriculture to address the association, reported that previous engagements would prevent the Secretary from accepting but that the Assistant Secretary hoped to be at the meeting Wednesday morning. Later engagements prevented the Assistant Secretary also from speaking. The following resolutions, offered by E. F. Ladd, were passed by the association : Resolved, That the president and executive committee of this association memo- rialize the President and Congress of the United States to enact legislation author- izing the Secretary of Agriculture to establish definitions for the better enforcement of the food and drugs act of June 30, 1906. Resolved, That the president and executive committee of the association be instructed and authorized to appoint a committee of three on definitions to codper- ate with a like committee of three already appointed by the American Food and Drug Commissioners. The members of the committee were announced later as follows: Wil- liam Frear, State College, Pa., Julius Hortvet, St. Paul, Minn., and J. P. Street, New Haven, Conn. R. W. Thatcher made the suggestion that a ten-minute limit be placed on the reading of all papers thereafter during the convention and the chair so ordered. R. N. Brackett reported for the executive committee the following resolution : Resolved, That the incoming president shall appoint a committee of three to contract with a responsible publisher to publish the Proceedings of the association; a definite number of copies to be contracted for and to be furnished to the Association at a reduced price. This resolution was adopted by the association with the following supplementary motion by William Frear: Resolved, That should the said committee of publications find the method out- lined by the executive committee to be impracticable, they be further authorized to adopt such other means as in their judgment shall be best for the accomplish- ment of the end. 169 170 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 REPORT ON MEAT AND FISH. By W. B. Smiru, Associate Referee. STARCH. The methods offered to the association last year were studied compara- tively this year. They are as follows: PRICE METHOD. Treat in a 200 ce. beaker 10 grams of finely-divided meat with 75 ce. of an 8 per cent solution of potassium hydroxid in 95 per cent alcohol, and heat on the steam bath until all the meat is dissolved. This will take from 30 to 45 minutes. Add an equal volume of 95 per cent alcohol, cool, and allow to stand at least 1 hour. Filter by suction through a thin layer of asbestos in a Gooch crucible or through an alun- dum crucible. Wash twice with warm 4 per cent potassium hydroxid in 50 per cent alcohol and then twice with warm 50 per cent aleohol. Endeavor to retain as much of the precipitate as possible in the beaker until the last washing. Place the crucible with contents in the original beaker, add 40 ce. of water and 25 ec. of concentrated sulphuric acid. Stir during the addition of the acid and see that the acid comes in contact with all the precipitate. Allow to stand about5 minutes, add 40 ce. of water, and heat just to boiling, stirring constantly. Transfer the solution to a 500 ec. graduated flask, add 2 ce. of a 20 per cent aqueous solution of phosphotungstie acid, allow to cool to room temperature, and make up to mark with distilled water. Fil- ter through a starch-free paper and after neutralizing determine the dextrose pres- ent in a 50 ce. portion of the filtrate with Fehling’s solution, using Low’s method, Bulletin 107, Revised, page 241, for the determination of the copper in the cuprous oxid precipitate. Dextrose X 0.9 = starch. MAYRHOFER-SACHSSE METHOD OF E. M. BAILEY. Treat 20 grams of the sample in a casserole with 50 cc. of 8 per cent solution of potassium hydroxid and digest on a steam bath with frequent stirring until the meat is entirely dissolved. Maintain the original volume approximately by the addition of water in case the solution concentrates appreciably. Add an equal volume of 95 per cent alcohol, stir thoroughly and filter on a small Biichner funnel provided with a thin asbestos mat. Should the filter clog and become slow, trans- fer the mat with residue to the casserole, removing any adhering residue in the funnel by means of a wisp of asbestos. Prepare a new filter. Add warm 4 per cent potassium hydroxid in 50 per cent alcohol to the residue in the casserole, disintegrate the mass by stirring and pour upon the new filter. Wash twice with warm 4 per cent alcoholic potash and finally with 50 per cent alcohol. (A new mat is not always necessary, but on the whole it wastes no time as the second filtration is always rapid. Moreover, the starch becomes incorporated with the asbestos and is better acted upon in the hydrolysis). Remove the filter to a 500 ce. flask, add 200 ce. of water, 20 ec. hydrochloric acid (specific gravity 1.125), and hydrolyze in a boiling water bath for two and one-half hours.. Nearly neutralize with dilute sodium hydroxid, cool, and make up to 500 ce. Determine dextrose in 25 cc. of the filtrate (equivalent to 1 gram of material), using Allihn’s copper solution. Multiply the weight of dextrose by 0.9 to obtain the weight of starch. The chemists who reported results on this work did some very thorough testing of the methods. Their reports follow: 1915] SMITH: MEAT AND FISH 171 REPORTS OF ANALYSTS. E. A. Boyer, South Omaha, Nebr.: A sample of cornstarch designated as c.p. was used in the determinations. The anhydrous substance was found to contain 90.5 per cent of starch by the diastase method. From a closed weighing flask and weighing by difference, quantities of the starch were added to chopped fresh lean beef. Comparison of results by methods for starch (Boyer). PRICE METHOD MAYRHOFER-SACHSSE METHOD Starch added Starch found Starch added Starch found per cent per cent per cent per cent 1.48 1.46 1.79 1.65 2.44 2.47 1.94 koe, 4.04 3.91 3.62 3.31 4.18 4.08 3.83 3.39 4.84 4.82 4.85 4.82 5.54 5.35 6.80 5.95 Since the fundamental principles of the two methods are nearly identical, results of equal value would be expected, but for possible errors in the manipulation. With the Mayrhofer-Sachsse method trouble was experienced in the filtering and the removal of the last traces of gelatinous starch from the casserole proved extremely difficult and time-consuming. The somewhat erratic results obtained by the method are attributed to these sources. As indicated by the tabulated data the Price method proved to be an extremely accurate one forthe purpose, and no difficulty was presented in the manipulation. The rapid means of digestion, lack of necessity of transferring all the starch precipitate to the filter, and the short time required for hydrolyzation are commendable points in the method. Regarding the determi- nation of dextrose, it would appear that this might be left optional, as results here obtained have shown that the cuprous oxid precipitate is nearly free from both organic and mineral matter, and, therefore, the gravimetric methods of weighing either as cuprous oxid or cupric oxid give results nearly identical with the volu- metric Low method. In laboratories making the determination frequently, possi- bly the Low method is preferable; if only an occasional determination of this kind is required, however, the making and standardization of the reagents necessary for conducting the volumetric method are to be considered. Using the Price method the following figures are offered from a number showing concordant results, to show results obtained by weighing the copper either as cu- prous oxid or cupric oxid, in comparison with estimating the copper by the Low method. Starch results using cuprous and cupric oxids and Low method. PER CENT STARCH SAMPLE Cuprous oxid Cuprie oxid Low method Meat food product........ 4.04 3.99 3.99 do 2.54 Py 2.51 do 3.70 3.69 3.69 do 2.96 2.92 2.92 Corn Holnsese sr eee 70.09 69.67 69.89 do OAs 70.10 69.90 do 66.85 66.85 66.90 172 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. J, No. 2 W. B. Meyer, Washington, D. C.: A comparison of the two methods shows that the Price method is much simpler in manipulation, is less liable to error, consumes less time, and gives more accurate results than the Mayrhofer-Sachsse method. With the latter, using aqueous potash to dissolve the meat, the starch is precipitated in a form that is bulky and hard to filter, a single sample requiring sometimes as much as 3 hours in spite of all the precautions suggested in the method. Error is likely to creep in when this precipitate must be transferred to a flask and difficulty is sometimes experienced in washing the precipitate from the sides of the beaker and funnel. The length of time taken for hydrolysis, 2} hours, is another objection to the method. The Price method permits filtering the starch on an average of 10 minutes per sample. The precipitate is compact and may be filtered through a Gooch crucible, which may be put directly into the beaker which is used for digestion, and the hydrolysis conducted in about 20 minutes’ time without a single transference. Determination of starch (Meyer). PRICE METHOD MAYRHOFER-SACHSSE METHOD Starch added Starch found Starch added Starch found per cent per cent per cent per cent 4.01 3.96 1.01 1.27 8.01 7.88 2.50 2.73 10.02 9.68 4.51 4.82 7.28 7.04 6.16 6.50 8.62 8.56 5.75 6.25 5.00 4.89 6.00 6.42 4.00 3.97 2.60 2.60 3.00 3.05 2.50 2.60 3.20 3.24 3.55 3.44 4.20 4.10 2.85 2.66 6.20 6.18 0.36 0.10 ToC 0.75 0.66 J.C. Himes, Washington, D. C.: The Price method is far superior to the Mayr- hofer-Sachsse method from the standpoint of time consumed in making analyses, and the work is not tedious with the former, as it is with the latter method. I found it almost impossible to do the filtering required by the Mayrhofer-Sachsse method. Determination of starch (Himes). PRICE METHOD MAYRHOFER-SACHSSE METHOD Starch added Starch found Starch added Starch found per cent per cent per cent per cent 2.00 2.17 2.01 1.96 5.00 5.07 1.98 2.00 8.20 UME 1.90 1.84 10.00 9.90 5.04 5.18 9.00 8.74 3.10 3.40 1.00 0.86 2.60 2.64 2.08 2.12 9.00 8.95 5.03 4.99 0.50 0.38 10.00 9.93 1.03 0.71 0.50 0.50 1.75 1.50 bee she 1.24 0.93 2.50 1.89 1915] SMITH: MEAT AND FISH 173 W. B. Smith, Kansas City, Mo.: The methods being so much alike it becomes a question chiefly of ease of operation. A sample of sausage gave the following results: Starch Price method Mayrhofer-Sachsse method Per cent Per cent tit 2.65 2.09 2.63 The lower figures in the Mayrhofer-Sachsse method are probably due to loss of starch. This method is far more tedious and hardly as accurate as the other. CONCLUSIONS OF ASSOCIATE REFEREE. There is little to add to the statements of the analysts quoted. Both methods depend on the same principles, but the Price method is superior in point of time, in manipulation, and in accuracy. In the three series. of tests given above the difference between starch added and found are as follows: Price method Mayrhofer-Sachsse method Maximum..........0.26 0.85 Minimum.......... 0.00 0.00 AVELAZE concen. 200-10 0.23 The Price method is being used in the routine examination of thou- sands of samples of meat food products with unvarying success. Dupli- cates are found to run very closely. AMMONITACAL NITROGEN. The Folin method for ammonia in meats was approved last year for final action in 1913. The accuracy of the method was unquestioned, but the time and apparatus required under the form of the method used, which was that employed by M. E. Pennington for chicken flesh, and the foaming in some instances, made further study desirable. There was also approved for study a method of distillation in alcoholic vapor. During the year, however, Folin sent the associate referee an improved form of his method, which has since been published in the Journal of Industrial and Engineering Chemistry. This revision is par- ticularly adapted to meats, and while little codperative work has been done on it, the work reported last year covered the subject so thoroughly that there seems no reason why the revised method should not be adopted. The principles on which revision has depended are (1) the volume of the sample should be as small as possible in order to complete the oper- ation (evolution of ammonia) quickly; (2) the sample solution should be saturated for the same reason (this is accomplished by the use of potas- sium oxalate); (3) the foaming complained of may be obviated, as was originally stated by Folin in 1903, by adding heavy petroleum. 174 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 The associate referee obtained the following results with the Folin- Pennington method (as recommended last year), the alcoholic vapor method, and the new Folin method: Comparison of results by methods for ammoniacal nitrogen. AMMONIACAL NITROGEN SAMPLE Alcoholic vapor Folin-Pennington New Folin method method method per cent per cent per cent is 0.0105 Potted meat........ OlotorGE WAT 4 Vana {0.0104 Meat extract....... 0.24400 OF25400) i ieee ak 0.2300 2 0.2300 do OP28000 Ae es eee 0.2100 oe 0.2200 RSE Oe { 0.07700 02040009 9h) aaa ea 0.01700 Witenebonsapesaugas > {0.02100 O:00100)0 | tia ae aes 0.0045 0.00280 URES» «ci Jee ene sels 0.00410 01002707 tlt eee do 0.00182 OXO0078;) 7 Ni eee The samples of meat and meat extract show that all three methods are accurate with these substances, but for two classes of substances, such as peptone and urea, which contain large proportions of compounds easily broken down to ammonia, and such as eggs, in which the quantity of ammonia is so small that the errors in titration and of hydrolyzation assume large proportions, the alcoholic vapor method is much less accu- rate than the aeration methods. And since the latest modification of the Folin method makes it as easy of operation as the alcoholic method, the Folin method seems the best possible. The method as stated below is delicate enough for ordinary purposes, but if greater sensitiveness is re- quired a larger sample can be used. The apparatus shown herewith is not large enough for eggs. The methods used are as follows: ALCOHOLIC VAPOR METHOD. Mix 25 grams of the fine sample with 5 grams of salt and 1 gram of sodium carbon- ate and 100 ce. of alcohol in a 500 ce. flask and pass through it a current of vapor from boiling alcohol. Distill a 200 ce. portion and titrate and then treat similarly two 100 ce. portions. The last portion should contain very little ammonia. FOLIN-AERATION METHOD. Arrange 5 vessels in a series as follows: (1) A bottle containing sulphuric acid with a Hopkins safety bulb, to purify the entering air; (2) a 1,000 cc. flask containing 25 grams of sample, 250 cc. of water, 5 grams of sodium chlorid, and 1 gram of sodium carbonate; alcohol may be added to prevent foaming; (3) a 250 ec. 1915] SMITH: MEAT AND FISH 175 safety flask; (4) a cylinder, fitted with a Folin absorption tube, containing tenth- normal sulphuric acid; (5) a 100 ce. safety flask. Connect the last flask to an air pump powerful enough to draw the ammonia over into the standard acid. Alcohol may be substituted almost wholly for the water if the air current is weak. Titrate the standard acid at intervals of an hour until no more ammonia is given off. Run a blank at the same time. Methyl Red, Cochineal, or Congo Red may be used in aqueous solutions, Methyl Red or Cochineal in alcoholic. NEW FOLIN METHOD. Weigh out 20 grams of finely-divided meat in a 50 cc. flask; add 10 cc. of water and 10 ce. of normal hydrochloric acid; let stand a few moments with occasional shaking; — > Sycfie7" Air From Wash gort/e. EON Ri acid: FIG. 1. APPARATUS FOR FOLIN AMMONIA METHOD. make up to mark, and shake. Allow solid particles to settle as much as possible, pipette out 10 cc. and place in Folin apparatus shown in accompanying sketch. Place twenty-fifth-normal or fiftieth-normal acid in receiving tube and then add to the sample tube, 1 cc. of saturated potassium oxalate, a few drops of kerosene, and finally 2 ec., or enough to make alkaline, of saturated potassium carbonate. 176 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 Immediately close the apparatus and pass air, washed in sulphuric acid, through it. Collect the ammonia in the standard acid and determine its amount by titration or by Nessler’s reagent. COMMENTS OF ANALYSTS. F.C. Weber, Washington, D. C.: We are at present using the Folin method as you describe it and consider it the best method. F. C. Cook, Washington, D. C.: I have had some experience with the different methods and consider that the Folin method, where the amount of ammonia is determined by nesslerization preferably, and by titration, is undoubtedly the best method available at the present time for the determination of ammonia. SUGAR. The association voted in 1909 that the referee give special attention to the apparent error in the method for the determination of sugar in meat and meat products, noted by Lowenstein in the Journal of the American Chemical Society, September, 1908. This error was said to be due to the presence in the 500 ec. to which the sample was first diluted, of the insoluble portion of the 100 grams of meat taken. But the insoluble solids of meat are not enough to account for the error found. Then in well over a hundred determinations the greatest plus error ever found by the writer was 0.10 per cent, that is, 3.71 per cent were found when only 3.61 per cent were added. This may well be due to the said cause, but as in practically all cases results are low the error does not seem to be large enough to affect the results. Furthermore, the same results were obtained, whether the sugar was added before separating the solid portion or whether it was added to the filtered juice. Further discussion of this point will therefore be omitted, leaving the figures to be given to speak for themselves. It was found, however, that the provisional method was not satisfactory for other reasons. In ordinary meats the percentage of sugar, both reducing and cane, is very small; hence it was thought necessary that a method should be found accurate to at least 0.05 per cent. The best method found in the literature appeared to be clarification with mercuric nitrate according to the description in C. A. Browne’s “Handbook of Sugar Analysis,’ and accordingly the later work has been an effort to put this method into suitable form for meats. Mercuric nitrate alone does not remove enough of the nitrogenous bodies; for that reason phos- photungstie acid has been added with considerable success. CLARIFICATION OF THE SUGAR SOLUTION. Three methods were studied: (1) Provisional (Bur. Chem. Bul. 107, Rev., p. 111).—Boil 100 grams of the fine- ly-divided meat for 15 or 20 minutes in a 500 cc. graduated flask, with a convenient 1915| SMITH: MEAT AND FISH W7iz volume of water. Add a few cubic centimeters of normal lead acetate, cool to room temperature, make up to mark with water, and filter through a folded filter. Re- move the lead and determine reducing sugar as dextrose, as described under ‘‘VI General Methods,” page 49. (2) Same with the further addition of phosphotungstic acid. (3) Boil 100 grams of finely-divided sample for 20 minutes with 350 ec. of water, cool, add 6 to 10 ec. of mercuric nitrate solution,! make to 500 ce. exclusive of fat (a graduate is convenient), and filter. To each 100 ce. of the filtrate taken, add 1 to 2 ec. of concentrated hydrochloric acid, heat to boiling, saturate with hydrogen sulphid, remove hydrogen sulphid by a current of air, cool, add 2 ce. of 20 per cent solution of phosphotungstic acid to each 100 ce., make to same volume as taken, and filter. Comparative results using different clarification methods, the reduction methods being comparable. REDUCING SUGAR AS DEXTROSE SAMPLE ee eae ak Calculated | Found by Found by Found by present Method 1 Method 2 Method 3 per cent per cent per cent per cent SIEEY EE SO ere 2.17 1.84 1.92 2.04 ere 1.91 1.93 2.03 SNe eS 2.11 AB ws sale Fete 2.02 do PAN 1.96 2.08 PPP. an oe eee 2.28 do 4.81 4.82 ait was Cured meat..:..... 8.00 7.90 aeta do 4.00 4.04 Wee do 4.00 3.62 sens Potted beef........ 3.61 3.69 3.71 Dried beef........ 2.61 2.29 2252 Cured beef......... 1.50 1.45 i G8} re NRUSAP Cuan: 1.02 0.85 oe 0.95 Cured ox tongue... . 0.92 dhe 0.84 ee Sausapes.cn eee 1.00 0.61 zoey Cured ox tongue... 0.50 0.27 0.35 While in the foregoing table the results by Methods 1 and 2 are in many cases good, with low percentages of sugar this is not so, the reduced copper often being held in solution or suspension by the nitrogenous bodies present, so that no precipitate is formed. The treatment with lead acetate, moreover, is not satisfactory, filtration being difficult and tedious. Before taking up Method 3, the copper reduction will be discussed. Only gravimetric methods have been studied, volumetric methods not appearing to have any superior advantages. Munson and Walker’s ‘ Mercuric nitrate solution: Treat 220 grams of yellow oxid of mercury in 300 cc. of water with small portions of nitric acid, warm, and stir until dissolved. Make to 1,000 cc. and filter. 178 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 method, being official, has been chiefly used. This method is not so accurate when small amounts of sugar are present as with large amounts, a fact true of Allihn’s also. Pfliiger and Koch and Ruhsam, offer modi- fications of Allihn’s method, chiefly consisting in boiling for 30 minutes, to overcome this, using, of course, a different table of copper values. In cases where the cuprous oxid was precipitated clearly and red the usual Munson and Walker method was followed, the copper being de- termined by Low’s iodid method. It sometimes happens, however, that the nitrogenous bodies still remaining in the sample solution, interfere with the reaction so that the copper is a muddy yellow. It was found that the following procedure gave practically the same ratios of copper found to sugar added as given in Munson and Walker’s tables (in the columns ‘‘Copper” and ‘ Dextrose’’): Modified Munson-Walker method of reduction.—Neutralize 50 ce. of the filtrate from the phosphotungstic acid (Method 3) with concentrated sodium hydroxid and place in a 400 ec. beaker with 25 cc. each of Soxhlet’s solutions, copper and tartrate (Bur. Chem. Bul. 107, Rev., p. 42). Bring to a boil in 4 minutes and boil 2 minutes. Filter through an alundum crucible of suitable porosity (the associate referee uses some numbered 5204-RA-360) with very gentle suction. If the filtrate is green or yellow it must be refiltered until clear blue. Wash the residue with a very little 5 per cent sodium hydroxid, refiltering the washings if they are muddy. Dissolve the residue in 1 to 1 nitric acid and determine the copper by Low’s iodid method (Bur. Chem. Bul. 107, Rev., p. 241). The treatment with bromin is unnecessary. The following table shows results obtained by clarification according to Method 3 and reduction as just described, except that in some of the early experiments variations occurred. Dextrose was used in all except Experiments 4 and 5, in which cane sugar was employed. Quantitative results were obtained in both cases, showing that hydrolysis of cane sugar takes place. The lowest results given are in Experiment 2, in which too little mercuric nitrate was employed. The difficulty of determining the blank correctly, the figure usually being low, may account for more sugar being found in some cases than was added. The average error of the above results is 0.02 per cent. Later work shows that the maximum error need not be more than 0.03 per cent. If not allowed to stand too long before filtering off the mer- curic-nitrate precipitate even large amounts of starch do not cause a serious error. Since writing the preceding part of the report comparative tests have shown that the Bertrand method of reduction, when the solutions have been properly clarified, gives the red cuprous oxid uniformly, also that the results are much less affected by the impurities present than by the 1915] SMITH: MEAT AND FISH Sugar in meat by method 3. SAMPLE 1—Fresh beef......... 2— do HAV OT ALO f: accysiaisisyacss seers IAVOT RDO oc aisle creer 3—Corned beef......... 4—Summer sausage..... 5—Fresh beef......... 6—Sausage............ 7—Cured meat........ 8—Sausage............. 9—Fresh beef......... 10—Corned beef........ 1—Sausage!........... 2—Fresh beef!.......... REAGENT TO 500 cc. OF SAMPLE : Phospho- M \ nitrate | *ngstic cc. cc. 5 Some 6 10 6 15 6 20 6 25 6 30 6 35 6 10 6 10 10 10 10 10 10 10 6 10 6 10 5 0 Some Some ‘5 10 6 10 10 10 10 10 SUGAR, BLANK ON MEAT per cent TOTAL 179 5 SUGAR appep |STGAR CAL-| pounp per cent per cent per cent 3.00 3.00 2.92 Bas Rae 2.98 0.98 1.05 0.99 0.98 1.05 0.94 0.98 1.05 0.98 aayate Sans 0.97 0.98 1.05 0.94 0.98 1.05 0.93 0.98 1.05 0.95 dno acer 0.94 0.53 0.66 0.63 UBas 1560 0.67 0.61 0.64 0.67 0.63 Sar yen 0.66 0.53 0.62 0.61 food eptele 0.71 ate te 0.62 430 5006 0.57 ed Foe Sitio 0.62 0.30 0.44 0.43 0.40 0.54 0.56 Fobc aie 0.55 ae ete oes 0.52 0.50 0.64 0.65 0.40 0.57 0.56 0.40 0.47 0.46 0.52 0.52 0.54 cone 560g 0.52 Bert bod 0.53 Bree Rh ee 0.53 0.05 0.13 0.11 0.10 0.18 0.17 0.25 0.33 0.32 0.10 0.10 0.08 ee eee 0.12 0.50 0.53 0.54 0.25 0.31 0.38 0.50 0.56 0.59 1 Results of C. T. Alleutt. Munson and Walker method. One cause of this is the greater dilution of the solutions. CLARIFICATION WITH MERCURY AND PHOSPHOTUNGSTIC ACID ACCORDING TO BERTRAND. ' The method which has given best results is as follows: AND REDUCTION Boil 100 grams of sample with about 350 cc. of water for about 20 minutes, cool, add an excess (10 to 30 ce. of twice-normal solution) of mercuric nitrate or acetate, 180 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 nearly neutralize with sodium hydroxid, and make to 500 cc. exclusive of fat. Allow to settle and decant the clear liquid through a large folded filter without washing. Make a measured volume of the filtrate sufficiently acid (1 or 2 ce. of concentrated hydrochloric acid per 100 cc.), heat to boiling, thoroughly saturate with a rapid current of hydrogen sulphid, remove excess of hydrogen sulphid by a current of air, make to volume, and filter. Adda slight excess of 20 per cent phosphotungstic acid (in 2.5 per cent hydrochloric acid) to the filtrate, allowing for the extra volume, and place in ice-box overnight or several hours at least. Filter and determine dextrose by Bertrand’s method as follows: Make the copper sulphate solution of 40 grams to the liter, and the tartrate solution, 200 grams of Rochelle salts and 150 grams of sodium hydroxid per liter. Neutralize 20 ec. of sample with sodium hydroxid, add 20 cc. each of the copper and tartrate solutions, in a 150 ec. Erlenmeyer flask; heat to boiling and boil gently 3 minutes. Filter off the cuprous oxid, wash with water and determine copper by Low’s iodid method. A part of Bertrand’s copper-dextrose table is as follows: Mo. of Mg. of Mg. of Mg.o Mg. of Mg. of Mg. of Mg. of copper sugar copper sugar copper sugar copper sugar 20.4 10 36.2 18 51.5 26 66.5 34 22.4 11 38.1 19 53.4 27 68.3 35 24.3 12 40.1 20 55.3 28 70.1 36 26.3 13 42.0 21 57.2 29 72.0 37 28.3 14 43.9 22 59.1 30 73.8 38 30.2 15 45.8 23 60.9 31 75.7 39 32.2 16 47.7 24 62.8 32 10.5 40 34.2 17 49.6 25 64.6 33 79.3 41 If the amount of copper found is less than 20.4 mg., the sugar value may be ap- proximately calculated, or the sample may be concentrated on the water bath after neutralizing with sodium hydroxid and acidifying again with acetic acid. The former method is the easier and apparently as accurate. RECOMMENDATIONS. It is recommended— (1) That under XVII, Methods for the Analysis of Meat and Meat Products on page 106 of Bulletin 107, Revised, 1, Identification of Species, 4th line, after ‘“melting-point,’”’ ‘melting-point of stearin by Belfield- Emery method,” be inserted. (This was adopted last year for the first time.) (2) That the Price method be made the official method for starch in meat food products, in place of Mayrhofer’s method, modified, 8, (b), (2), page 109, Bulletin 107, Revised. (3) That the new Folin aeration method (J. Biol. Chem., 1912, 11: 493, 523, 527; J. Ind. Eng. Chem., 1913, 5: 485) be introduced as (g) Ammonia, on page 109, Bulletin 107, Revised. (4) That under (e) Ammonia, page 115, Bulletin 107, Revised, the following be substituted: ‘‘Mix 1 gram of meat extract with 2 ec. of normal hydrochloric acid and wash into the Folin apparatus with about 5 ec. of water; proceed as under (g) Ammonia, page 109.”’ 1915) KERR: FATS AND OILS 181 (5) That the mercury-phosphotungstic acid method for clarifying meat sugar solutions, with reduction according to Munson-Walker or Bertrand, be studied further. REPORT ON FATS AND OILS. By R. H. Kerr, Associate Referee. GLYCEROL SAPONIFICATION METHOD FOR THE TITER TEST. For the work on the glycerol saponification method for the titer test, recommended by the associate referee last year, the following instructions were sent to the collaborators selected for this part of the work: INSTRUCTIONS TO COLLABORATORS. Three samples are being sent you, one of cottonseed oil, one of inedible grease, and one of oleostearin. It is requested that you determine the titer of these samples by the glycerol method and by the official method as given in Bulletin 107, Revised, page 135. Glycerol method—Heat to 150°C. in an 800 ce. beaker, 75 cc. of a glycerol potas- sium hydroxid solution, made by dissolving 25 grams of potassium hydroxid in 100 ce. of high test glycerol; then add 50 cc. of the oil or melted fat, previously filtered, if necessary, to remove foreign substances. Saponification in many cases takes place almost immediately, but heating, with frequent stirring, should be continued for 15 minutes, care being taken that the temperature does not rise much above 150°C. When the saponification is complete, as indicated by the perfectly homogeneous solution, pour the soap into an 800 cc. casserole containing about 500 cc. of nearly boiling water, carefully add 50 cc. of 30 per cent sulphuric acid, and heat the solution, with frequent stirring until the layer of fatty acids separates out perfectly clear. Transfer the fatty acids to a tall separatory funnel, wash three or four times with boiling water to remove all mineral acids, draw them off into a small beaker, and allow to stand on a steam bath until the water has settled out and they are clear. Filter into a dry beaker and heat to 150°C. on a thin asbestos plate, stirring con- tinually with the thermometer, transfer to a titer tube, filling it to within 1 inch of the top, and take the titer as indicated in the present provisional method. You are also requested to determine the iodin number of the fatty acids obtained by each method. RESULTS OF COOPERATIVE WORK. The results reported by the collaborators are given in the following table: 182 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 Report of collaborators on methods for making the titer test. GLYCEROL SAPONIFICATION PRESENT PROVISONAL METHOD METHOD SUBSTANCE AND COLLABORATOR a ee AG: °C. Cottonseed oil: L. B. Burnett (Bureau of Chem- "GDA fetaqanne stan SOG Snioser SASS) Mas by clerics SL ogee eee 34.9 TTA 55 34.8 110.8 H. C. Fuiler (Institute of In- dustrial Research)........... Way arf 78.00 132.85 83.00 R. R. Henley (Bureau of Ani- malindustuy) menses s eee 34.35 105.06 34.3 106.92 R. H. Kerr (Bureau of Animal Indus tiny) haeeocre eee eee 34.1 107 .60 34.1 107.70 Paul Rudnick (Armour and Co). 34.4 110.75 34.0 107.55 F. N. Smalley (Southern Cot- POMOC O)jae ese eee 34.2 102.20 35.0 105.73 Maximilminns te tesnr eerie cee SEO aL eee 3D. ||| hneee IMI pee Geka es Oe eee Sy Tt ee |e Goencteieys 34:0). || = *ae eee AW EAGER. scene ee cers cere aero ae be: Miah I oe ses B45 NS eee Inedible grease: Ey yBeeBurnettsc.< see cele ac 41.5 as ois 6 pegs 41.6 57.2 41.5 BY 6H Him Corbullerte eect pice ivcs ace 140.0 33.0 140.0 29.0 Re RS clenleyil seek ents 41.45 55.08 41.55 56.07 FR resKern iwspem tc cise eee 41.2 54.88 41.2 54.34 RaulsRidnickt hs seek acetals 41.5 56.27 41.3 56.20 Ho INeiSmalleys-ce. cotta eee 41.6 50.76 41.7 53.15 INGER aTN ssa Red op oo me Glune oot AA GME Ailey Peon ee a7" | ee Mani UM Seis eee eee Oil aPhee (pee! bderstie 41.2 ~)| \ ee IAVOTARS Pers dee Net he es aed ANAT pete) 's dan Seek cre 41.45, «is Coes Oleo stearin: ieaBeoburneuoeteatcetces cen dee SO Se eee 50/6:) «1)" Ea 50.9 24.0 50.7 23.2 Ey Csabullernswe ye a Fas cares IO Sy leu. seek 149 95) ‘inde Ween RSS eenle yi ise a: etree ose sivas 50.65 21.79 50.7 22.90 Re IRerrep pies series oon 50.8 23.40 50.8 24.34 Pauli Rud ni cksyenaeee cine meee 50.9 23.78 50.8 22.93 IBSING i Smalleyaeess. cee cesar 50.8 21.68 Sie 22.27 [Miaximinare seen. cte nceeek cee SOLO ASIF Poa -cee 5b. 1. pe eee Minimumin secant ice ee GO)-Gite |B YO aeae 5OLG2) Alle seeeee PA OT SD Says aise aera ads waren eis OSS | meer otctcr 50:8". | = ees 1 Not included in average. COMMENTS. The results appear to give the glycerol method a slight advantage over the present provisional method. In each case the maximum titer obtained by the glycerol method was lower and the minimum higher than by the present method. The averages were identical in two cases out of the three. Paul Rudnick: Our results show that a slightly higher titer is consistently ob- tained by the glycerol method, this difference ranging inversely as the hardness of the fat. We are inclined to lay this to two features, one the use of alcohol in the 1915] KERR: FATS AND OILS 183 official method, the other the difference in the method of drying. When alcohol is used it is necessary to get rid of the last traces so as not to lower the titer of the resulting fatty acids. This commonly induces an error in the opposite direction, namely, scorching of the fat. Both dangers are entirely avoided by the glycerol method. L. B. Burnett: Have obtained practically identical results with the two methods. Prefer the glycerol method for the following reasons: It is easier to manipulate, it is quicker, and the fatty acids are not so difficult to decompose when glycerol and potash are used for saponification. As the results show clearly that the glycerol saponification method is accurate and reliable, and it is beyond question more rapid and conven- ient than the present method, I recommend that it be adopted as an offi- cial method. EMERY METHOD FOR THE DETECTION OF BEEF FAT IN LARD. A test was made of the Emery method with mixtures of lard and beef fat, lard and mutton fat, and lard and hydrogenated cottonseed oil. Hight samples were sent to the collaborators on this work, the composi- tion of these being as follows: Sample 1 Lard containing 5 per cent of beef tallow. Sample 2 Lard containing 3 per cent of mutton stearin. Sample 3 Lard containing 5 per cent of lard stearin. Sample 4 Pure lard, 65 per cent; hydrogenated cottonseed oil, 15 per cent; cottonseed oil, 20 per cent. Sample 5 Lard containing 2} per cent of oleo stearin. Sample 6 Pure lard. Sample 7 Lard containing 10 per cent of tallow. Sample 8 Lard containing 23 per cent of hydrogenated cottonseed oil. The following instructions were sent to the collaborators selected for this part of the work: INSTRUCTIONS TO COLLABORATORS. Eight samples are being sent you for the determination of the melting point of the stearins by the Emery method (Bureau of Animal Industry Cir. 132). You are also requested to give your judgment as to which of these samples are pure lard and which are adulterated. In reaching a decision as to the purity or otherwise of any particular sample, you need not depend wholly upon the melting point of the crystals, but may use any other factors you consider of value in reach- ing a decision. In reporting your results please state, first, melting point of stearin erystals and temperature at which crystallization took place; second, judgment as to the purity of each sample, that is, whether it is a pure lard or not; third, what factors, if any, other than the melting point of the stearin crystals were taken into account in forming your judgment as to the purity or otherwise of each sample. RESULTS OF COOPERATIVE WORK. The results reported by the collaborators are given in the following table: 184 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 Reports by collaborators on the Emery method for beef fat in lards. MELTING POINT OF CRYSTALS COLLABORATOR AND CRYSTALLIZED AT— TODIN SAMPLE NS BER 12 °C. 16 C. 18 °C. Sc: oC. eG. L. B. Burnett, Bureau of Chemistry: 62.0 62.2 62.6 61.4 62.0 62.5 62.4 61.5 62.4 63.3 64.0 63.4 62.6 64.0 64.2 64.0 62.4 62.5 62.4 62.2 63.4 63.8 64.4 63.5 60.8 61.8 61.6 60.7 (ened 61.2 61.9 63.0 62.5 H.S. Bailey, Bureau of Chemistry: 62.5 63.0 62.6 63.0 63.9 64.8 64.0 64.4 62.7 63.8 64.0 64.3 62.4 62.0 62.0 62.8 63.9 64.0 Yop H. E. Woodward, Bu- reau of Chemistry (Philadelphia): 61.9 62.2 61.8 61.7 63.1 63.2 62.6 62.9 62.0 61.9 63.0 63.0 61.0 61.3 61.6 61.3 Pure lard... ..| 63.4 63.7 Lard stearin......} 62.5 63.2 Tallow ..| 57.8 57.3 Oleo stearin.......| 57.6 57.4 16°-18°C. I. R. Howlett, Wiscon- sin Dairy and Food ROK Commission: 62.8 62.6 64.2 64.4 63.2 64.0 62.3 Chg sengnouserssbae 62.4 JUDGMENT Adulterated do Pure Suspicious Adulterated Pure Adulterated Pure Adulterated do do Pure (?) Pure do Adulterated do Pure Adulterated do Pure Adulterated do Pure Adulterated do Adulterated do Pure Adulterated do Pure Adulterated do OBSERVATIONS Remained at a temperature of 18°C. for 4 days before erystals were obtained. Appears suspicious but has correct melting point. Microscopic appearance of crystals; melting point of fat. Microscopic appearance of crystals. do 19165] KERR: FATS AND OILS 185 Reports by collaborators on the Emery method for beef fat in lards—continued. MELTING POINT OF CRYSTALS} | (HEME GH ES) CRYSTALLIZES NuM- | JUDGMENT OBSERVATIONS SAMPLE || 14 °C. | 16°-18°C.} 18 C. Ve sGs SCs °C. R. H. Kerr, Bureau of Animal Industry: Lee ciotsis niocieeiteciles dacs 00nd 63.4 oboe Adulterated Oh noposondoacaadce eaters asters 62.4 sete do Bh; oocagdoonoDEn ss: abc c1S05 64.0 9960 Pure ( WES BUR oooecn nen Tats ofatels 64.0 O500 Adulterated | Crystals while showing the melting point of lard crys- tals, are totally different in appearance. ocbeuacheEmeeaone onde s$c6 63.4 SOOE do (acmnascodsonscae soon sc00 64.2 S00 Pure ASG IOOROCERECACOD ayes Sano 61.8 od5e Adulterated Uoddariganosbocbsd ene Sate 62.8 anad do 15°C. 17°C. 19°C. *¢ °c °¢ W. B. Smith, Bureau of Animal Industry (Kansas City): 62.7 62.8 Saes Adulterated etateta 62.7 62.7 Riel do 63.9 64.0 64.1 nso Pure 64.0 64.1 64.1 Aries do Had such a heavy precipi- tate that it seemed im- possible that it could be pure but the Halphen test gave 20 per cent of cotton- seed oil so this sample seems to be a mixture of lard stearin and cottonseed oil. 62.7 62.9 ites pcon Adulterated 63.8 64.0 64.1 pone Pure 62.4 62.1 state} Adulterated 61.7 61.5 AnDC do 17°-18°C. °C: R. R. Henley, Bureau of Animal Industry: 63.2 S000 saa Adulterated 62.4 chee peo do 64.1 andr seas Pure 64,2 soap ent Adulterated Gave a large deposit of dull crystals totally different in appearance from those ob- tained in pure lard. 62.9 ayaiacs 2000 do 64.0 “006 oon Pure 62.0 Sco8 bot Adulterated 62.2 50an aceo do 186 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 Samples 1 (lard containing 5 per cent beef tallow), 2 (lard containing 3 per cent mutton stearin), and 7 (lard containing 10 per cent beef tallow) were pronounced adulterated by all collaborators. Sample 6 (pure lard) was pronounced pure by all. Samples 5 (lard containing 2% per cent oleo stearin) and 8 (lard containing 24 per cent hydrogenated oil) were judged pure lard by one collaborator and Sample 3 (lard containing 5 per cent lard stearin) was judged adulterated by one. Sample 4, whose adulter- ants were so chosen and proportioned that the crystals obtained would show the normal melting point of pure lard, was reported as suspicious or adulterated by all of the collaborators. This sample gave a large mass of dull crystals, totally different in microscopic appearance from those given by lard. These results show that the Emery method is not only a reliable method for the detection of beef fat in lard, but is also of value in the detection of other adulterants. The melting point of the crystals was depressed by all of the adulterants tested except the mixture of cottonseed oil and hydrogenated oil. In this case the lard was grossly adulterated without depressing the characteristic melting point of the stearin crystals, but its character and the character of the deposit were so changed that adul- teration was suspected by all of the collaborators. RECOMMENDATIONS. Tt is reeommended— (1) That the glycerol saponification method for the titer test as given in this report be adopted as official. (2) That the Emery method as given in Bureau of Animal Industry Circular 132 be adopted as a provisional method for the detection of beef fat and other solid fats in lard. REPORT ON DAIRY PRODUCTS. By Junrus Hortvet, Associate Referee. Following the recommendations adopted at the meeting in September, 1912, the work of the present year has been directed toward a study of the modifications of the continuous extraction method for determining fat in cream, homogenized cream, and ice cream. The work has included also comparative fat determinations by means of the Roese-Gottlieb method and an examination of the fat recovered by the continuous ex- traction method described by the associate referee in 1911. The collaborators were also requested to undertake a microscopic examination of the samples submitted with a view of the possible detec- tion of homogenization. One of the collaborators has given special attention to this feature of the work, and his report is submitted herewith. 19165) HORTVET: DAIRY PRODUCTS 187 Samples of various kinds of cream products, two of which were homog- enized and two of which contained known proportions of foreign fat, were submitted to the collaborators. The samples submitted included: (1) Pure untreated sweet cream; (2) same cream homogenized; (3) same cream homogenized with oleo oil (added to constitute 20 per cent of the fat); (4) ice cream made from same cream with gelatin, sugar, and cottonseed oil (added to constitute 20 per cent of the fat). All samples were preserved with a little formaldehyde. The plan of preparation of the samples for the fat determinations has been somewhat revised, although it differs in no essential respects from the usual procedure. The aim has been to adapt the method of prepa- ration to the character of the samples to be extracted. The collaborators were requested to make the following determinations upon the fat recovered from the various samples:—Reichert-Meissl number, iodin number, refractive index at 25°C. The methods for Reichert-Meiss] number and refractive index are in all essential respects as described in Bulletin 107, Revised, pages 131, 136, and 141. The method for the Reichert-Meiss] number determination is essentially the Leffman and Beam method given on page 141. The associate referee is indebted to E. H. Farrington and A. C. Baer, of the University of Wisconsin Agricultural Experiment Station, for their services in preparing these samples and forwarding them to the various collaborators. The entire description of the methods as submitted to the collaborators is given as follows: INSTRUCTIONS TO COLLABORATORS. DETERMINATION OF FAT. A—Roese-Gottlieb Method. (1) Milk (whole or skimmed).—If the milk be soured, obtain an even emulsion preferably by repeatedly pouring the sample back and forth from one container to another. Unless a fine, even emulsion can be secured, it will not be expected that a satisfactory analysis can be made. Weigh out 3 to 5 grams of the homogeneous sample and transfer to a Rohrig tube (Zts. Nahr. Genussm., 1905, 9: 531) or to a suitable size Werner-Schmidt extraction apparatus (Leach, 3d Ed., p. 139), using for the purpose not more than 10 cc. of water. (2) Evaporated milk (sweetened or unsweetened) and cream.—Mix the sample thoroughly, best by transferring the entire contents of the can or bottle to a large evaporating dish, and working it with a pestle until homogeneous throughout. Weigh 40 grams of the prepared sample, preferably in a tared weighing dish used for sugar analysis, transfer by washing to a 100 cc. graduated sugar-flask and make up to the mark with water. Measure into one of the extraction tubes described in the preceding paragraph 10 cc. of this solution. (3) Ice cream.—Soften the sample, by warming, to the consistency of ordinary cream, transfer to a beaker and stir until homogeneous throughout. Weigh out 188 | ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 3 to 5 grams of the sample, transfer to an extraction tube, using for the purpose not more than 10 ce. of water. (4) General procedure.—To the material in the extraction tube add 1.25 cc. of concentrated ammonium hydroxid (2 ce. if the sample be sour) and mix thoroughly. Add 10 ce. of 95 per cent alcohol and mix well; add 25 ec. of washed ethyl ether, shake vigorously for a half minute, add 25 cc. of petroleum ether (redistilled slowly at a temperature below 60°C. preferably) and shake again for a half minute. Let stand 20 minutes or until the upper liquid is practically clear and its lower level constant. Draw off the ether fat solution as much as possible (usually 0.5 to 0.9 cc. will be left) into a weighed flask through a small quick-acting filter. Re-extract the liquid re- maining in the tube, this time with only 15 ec. of each ether, shaking vigorously half a minute and allow to settle. Draw off the clear solution through the small filter into the same flask as before and wash the tip of the spigot, the funnel, and the filter with a few cubic centimeters of a mixture of the two ethers in equal parts. Extract again and wash in the manner just described. Evaporate the ether slowly on a steam bath, then dry the fat in a water oven until loss of weight ceases. B—Continuous Extraction Method of A. E. Paul. (1) Milk and evaporated milk (sweetened or unsweetened) and thin ice cream.—Pre- pare the material in the manner described under A. Into a 1,000 cc. beaker weigh 100 grams (in the case of milk 200 grams) of sample. Add 300 ce. of water, mix thoroughly and heat to boiling; add while boiling, very gradually 25 ce. of Soxh- let’s copper sulphate solution diluted with 100 ce. of water. In a Biichner funnel wet a filter of suitable size and of loose texture. Filter with suction and wash three times with a little boiling water, filtering as dry as possible. Remove the cake, which should be dry enough to be broken up easily between the fingers, break into small particles, and dry in the open air overnight. Grind in a mortar with sufficient (usually 25 grams) anhydrous copper sulphate, let stand a few minutes, or until the product seems quite dry and not at all lumpy. Into the inner tube of a large Soxhlet or other extraction tube place a layer of anhydrous copper sulphate, then the powdered mixture. Place on top a loose plug of cotton and extract 16 hours with ordinary ether. The ether should be poured into the extractor and allowed to percolate through before the heating is begun. About 50 cc. of the solvent will be required. Evaporate the ether slowly on a steam bath, then dry the fat in a water oven until loss of weight ceases. Reserve the weighed fat for further examination. (2) Cream and thick ice cream.—Prepare the material in the manner described under A. Ina Biichner funnel wet an 11 cm. filter of loose texture and cover with a layer of fibrous asbestos, being careful to cover the sides as far up as possible. In a 250 ce. beaker boil 25 ec. of Soxhlet’s copper sulphate solution, and add, while stirring vigorously, 50 grams of the material. Immediately remove the source of heat and filter with slow suction. Wash once or twice with a small amount of cold water, and proceed as in the method described for evaporated milk. EXAMINATION OF THE FAT, A—Reichert-Meissl Number. (1) Preparation of reagents —(a) Sulphuric acid—Dilute 200 cc. of the strongest acid to 1,000 cc. of water. (b) Barium hydroxid solution—Standardize an approximately tenth-normal solution. 1916] HORTVET: DAIRY PRODUCTS 189 (ce) Indicator—Dissolve 1 gram of phenolphthalein in 100 cc. of 95 per cent alcohol. (d) Pumice stone—Heat small pieces to a white heat, plunge in water, and keep under water until used. (e) Glycerol soda solution—Add 20 ec. of sodium hydroxid solution, prepared by dissolving 100 grams of sodium hydroxid, as free as possible from carbonates, in 100 cc. of water, to 180 cc. of pure concentrated glycerol. (2) Determination.—Add 20 ce. of the glycerol soda to approximately 5 grams of the fat in a flask, weighed accurately, and heat over a naked flame or hot asbestos plate until complete saponification takes place, as is shown by the mixture becom- ing perfectly clear. If foaming occurs, shake the flask gently. Add 135 cc. of re- cently boiled water, drop by drop at first, to prevent foaming, and 5 ce. of the dilute sulphuric acid solution, and distill, without previous melting of the fatty acids 100 cc. in about 30 minutes. Mix this distillate, filter through a dry filter, and titrate 100 cc. with the standard barium hydroxid solution, using 0.5 cc. of phenol- phthalein as indicator. The red color should remain unchanged for two or three minutes. Increase the number of cubic centimeters of tenth-normal alkali used by one-tenth, divide by the weight of fat taken, and multiply by 5 to obtain the Reichert-Meiss] number. Correct the result by the figure obtained in a blank experiment. B—Iodin Number. (1) Preparation of Reagents.—(a) Iodin solution—Dissolve 13.2 grams of pure iodin in 1,000 ce. of pure glacial acetic (99 per cent), and to the cold solution add 8 ce. of bromin, or sufficient to practically double the halogen content when titrated against the thiosulphate solution, but with the iodin slightly in excess. (b) Decinormal thiosulphate solution—Make by dissolving 24.6 grams of the freshly-powdered, chemically-pure salt in water, and make up to 1,000 ce. (ec) Starch paste—Prepare by boiling 1 gram of starch in 200 cc. of water for 10 minutes, then cool. (d) Potassium iodid solution—Make by dissolving 150 grams of the salt in water, and making up the volume to 1,000 cc. (e) Potassium bichromate solution—Make by dissolving 3.874 grams of chemi- cally pure potassium bichromate in distilled water, and making up the volume to 1,000 ce. (2) Standardization of the thiosulphate solution.—Introduce 20 cc. of the potassium bichromate solution into a glass-stoppered flask together with 10 cc. of potassium iodid and 5 cc. of strong hydrochloric acid. Then add slowly from a burette the sodium thiosulphate solution until the yellow color of the solution has nearly dis- appeared, after which add a little of the starch paste, and carefully continue the titration to just the point of disappearance of the blue color. Find the equivalent of 1 gram of iodin in terms of the thiosulphate solution by multiplying the number of cubic centimeters of the latter solution required for the titration by 5. The reciprocal of this result (expressed decimally) is the equivalent of 1 ec. of the thio- sulphate solution in terms of iodin. (3) Procedure.—Weigh out 0.3 to 0.5 gram of the fat in a glass-stoppered flask of 300 cc. capacity. Dissolve the fat in 10 ec. of chloroform, add 30 ce. of the iodin reagent, shake, and set in a dark place for half anhour. Add 10 ce. of the potassium iodid solution and 100 cc. of distilled water. Titrate the excess of iodin with the thiosulphate solution, which is slowly added from a burette until the yellow color has nearly disappeared, then add a little starch paste, and finally thiosulphate solution drop by drop until the blue color of the iodized starch is dispelled. Stopper 190 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 and shake the flask vigorously, and add sufficient thiosulphate to prevent a re- appearance of the blue color in five minutes. Conduct a blank determination at the same time, in exactly the same manner as in the above in order to obtain the cubic centimeters of thiosulphate solution equivalent to the 30 cc. of iodin solution used in the determination. Calculate the percentage of iodin absorbed (Leach, 3d ed., p. 489). C—Refractive Index at 25°C. Take at any convenient temperature and calculate to 25°C. by the formula: R= R + 0:55 (7’ — 7), in which R is the reading reduced to 7, R’ the reading taken at temperature T’ (Bulletin 107, Revised, p. 181). RESULTS OF COOPERATIVE WORK. The reports of the collaborators are included in the table on pages 186 and 187. DISCUSSION. All collaborators report difficulty in obtaining satisfactory results on Sample 1, owing to the somewhat churned or partly-separated condition of the fat. The other samples were received in good or fairly good con- dition. Allowing for these difficulties, however, the results are quite satisfactory, and the consensus of opinion is that the continuous ex- traction method applied to rich cream and cream products will not only yield reliable quantitative results, but a sufficient amount of fat in proper condition for chemical examination. The general trend of the results indicates that the new method recovers as a rule a larger proportion of fat, both in the unhomogenized and homogenized samples. There are some irregularities which can be accounted for by lack of experience on the part of the analyst. There is sufficient evidence in the results sub- mitted to ndicate that the modified new method may be relied upon when carried out under proper conditions and by an experienced analyst. The indications are that the results for fat will be higher than results obtained by the Roese-Gottlieb method. It is to be regretted that no reports have been received from two of ‘the persons who offered to collaborate, and that one or two others were not able to submit complete reports. It is shown by results that the Reichert-Meissl number is an important factor in judging the character and approximate percentage of foreign ‘fat. The refractive index figures are very sign ficant, especially in the case of Sample 4, and the iodin numbers on Samples 3 and 4 are decidedly corroborat've of indications afforded by the Reichert-Meissl determinations. The following comments and criticisms were made by the collaborators: “ 191 DAIRY PRODUCTS HORTVET 1915] egongqn0nnDog6 Soooo0t BORAT I ebeeooowad = |leduer reese boop T { dete) wig [ct unpeg 9 “5 Ce aeRO en leCoucE aan : { Be ee | ane Pe er ee [Io o8jo | ynoqy x Cy GF T91 0'F24 | oy-22 PIBMIOT “DO Be aes HERO seed RT OL lle tec Evra : { GOOF T SIGs ore aoe Rua cee { poe! | “FT lecort _| 6z'ez | 08°02 Tete ets ST "pezel 4u90 stereneiete ¢c’6e e¢ CT 0° #¢ 68°16 FOI Damp SN OS Ge sudas roqiuamiog } sodop-e¢ moay } (| See | Zeer jocort ja e( ote zearrpeao sd (0 saosoray | eae es San Nae | Rema shoes | otoee — || Saach CU | sao? || Hales eben, Oe --proyzaysatg. “HO :]I0 08]0 Y4TA poziue -souoy ‘T sev ureeIO ouIeg “g BS Pc Re tee |e nn ea ae eee } : souee Ihnen Fatcr eect feces wpe “9D 00'0F : ‘ 62° ST sb 000BDG0 Se sosdsdoosenos llonocusocan | 69° OF nee oe ope wat | doesgseneon ae pIrvaso,y “9 ‘O GO OF 68 ST EEE, eRe eee, een ey IL hes eee at zz { G29F T Ze GT feo een a. oe { 162¢$ | 29%) lpzopt | seer | or eT uigckoersle 2 lbs } : L8e | 04°8 9'e¢ | 9¢°ST aes bax ey, ey, 3 ee poop Say { ese joc |ztort | eget | tg°st pact eT), Si dope AP || lies eo hea ee ee one a [ge om reece 9°88 taza @o | eo St GPO Noonvessoogn ae { zee | ee escort | 6r eT eae Plogtors9rd “HO :peziue -dowoy ‘T sv weeds oUIeg *% Peery eects : Seaton lecaivsrntctels San Pct 20981 { pean ee Does see ued 49 49) Gr 8s : : ITT op | etter tteee foresee eee 0°68 ose) reat Signal os ae pawasoyg “Q ‘oO ; G8 FZ 0° ST pownyp ose £0' ST Fae eau perone node aataaeodn: { wyGs 09 124 Ze9r'T } re et { aes) See qaaumg “g “7 ee 69° 8E 1 lreor 1 IN | rea poyeiedes Alprg AON? ooreoreece { aes or re { Fegee | 89° ST f ee } SO kay IasuIpBuy “_ “O Adu] pue paumyg | ares ees hase { Ge LT tereeerssseng19aSOIG “FT * ozo | 9 ar poe FO eer ac (2 ewe Pley1}801g “H *O :]10 poos -u0}j00 puv iesns ‘ure -[03 YAM ‘T WoIy WALD DOT “p qu99 49d quao sad NOMWOVUL cdonLan ; AALOVULX ; wa AWN 009% : ; oe SS UVa NI AGhLVOIGNI Rebate rents nsaran! |llnvxnaN|leeierc.. ||| econ MOLYUOMVIIOD ANY WIA VS . DIG a are fs ~ \f ~ : IVa NOWMUOd ‘ DLUGHOLU | MAWLOV Uo AW Uva Aa Lva ‘ponulryuoj—suoynurusajap yof Wo s10zn.L0qQn}}09 fo sjLoday 1916) HORTVET: DAIRY PRODUCTS 193 COMMENTS OF COLLABORATORS. C. H. Biesterfeld, Bureau of Chemistry: Results by the Paul method are fine, being practically the same as by the Roese-Gottlieb. Whether the fats suffered any by standing a month in the refrigerator, I do not know. D. B. Gnadinger, Bureau of Chemistry (Chicago): It would seem that the continu- ous extraction method yields a quantitative recovery of fat, and that for most dairy products it has the advantage of calling for a larger quantity of fat, thus insuring a better average sample than can be secured by the Roese-Gottlieb method. In my estimation, therefore, the extraction method is satistactory for both purposes; the quantitative fat determination, and the separation of sufficient fat for exami- nation. Of the constants determined, it is my opinion that the Reichert-Meissl value is the one to be depended upon. The others may be useful for the purpose of corroboration.. As to the question of homogenization, nothing was done, except to note the striking difference in the condition of the homogenized and unhomogen- ized samples; the former were quite smooth, and any slight separation was easily stirred up, with the result that quite a homogeneous mixture was obtained, while the latter was entirely separated. L. B. Burnett, Bureau of Chemistry: Have obtained more consistent results with the continuous extraction method. While requiring much more time, no trouble was experienced with the method. The modification suggested for ice cream, that is, the manner of filtering on a pad of fibrous asbestos, although filtering slowly, a clear filtrate was obtained. Was considerably surprised to find that the per- centage of fat as obtained with the extraction method was invariably higher than that with the Roese-Gottlieb method. As a check the nonvolatile matter was estimated in the ether used for extracting, also the extract was dissolved in petroleum ether and weighed a second time, but the corrections introduced in this way were very slight. It is suggested that there was an incomplete extraction in the Rohrig tube. A third extraction was made in the Roese-Gottlieb method in addition to the two requested, but the amount of fat obtained was so small as to be practically negligible. C. C. Forward, Laboratory of the Inland Revenue Department, Ottawa, Can.: The results obtained by the method of A. E. Paul for fat in ice cream are not very satis- factory. If there had been more of the samples allowing for further work, results obtained might show to better advantage as the method requires some practice in manipulation in order to use it accurately. Sample 3 appears to be a mixture of butter fat with about one-third oleo oil, or about 73 per cent oleo oil in the total mixture. In Sample 4 about one-third of the fat appears to be cottonseed oil. G. G. Parkin, State Dairy and Food Department, St. Paul, Minn.: Some difficulty was experienced in grinding the dried material with anhydrous cupric sulphate until free from lumps. The large amount of oil or fat was probably the cause of this trouble. The tendency of the material to pack in the extraction tube afforded some trouble, which was overcome by the packing of two alternate layers of cotton and material. The continuous extraction method gave results which agreed very well with the Roese-Gottlieb method. Its application would be advisable in all cases where a large amount of oil or fat is desired for further analysis and where the time element is not to be considered. R.W. Hilts, Bureau of Chemistry (Seattle): I have made a microscopic exami- nation of the samples with a view of detecting homogenization. The samples were examined under a magnification of 450 diameters with a micrometer eye-piece. Sample 1: The fat globules vary much in size. The large ones range from 5 to 9.6 microns in diameter; the medium size, which predominate, range from 3.3 to 194 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 4.1 microns, and the smallest globules about 0.2 to 0.8 micron. In this normal sample, by far the bulk of the fat exists in the middle and large size globules, leaving considerable free space apparent between them. Sample 2: The large and medium size globules range from 2.7 to 6.9 microns in diameter and the small ones about 0.2 to 0.5 micron. In this sample the small globules predominate immensely in number, the medium and large ones being relatively few in number. The small globules are so numerous that free spaces be- tween them are very small. Sample 3: The large and medium giobules range from 1.6 to 9.6 microns in di- ameter, the small ones about 0.2 to1.3 microns. The statements made relative to Sample 2, as to general appearance and distribution of the globules, apply to this sample. There is noted in this sample, however, some very large globules that are even larger than those found in normal milk, due, doubtless, to the use of a foreign fat. Sample 4: The large and medium globules range from 2.7 to 8.2 microns in di- ameter and the small ones from 0.2 to 1.3 microns. The general appearance of the sample is the same as Sample 3. The use of a micrometer eye-piece is not necessary to distinguish the homog- enized from the natural product as the relative numbers of small, medium, and large globules are very striking to the eye, using the magnification stated. The measurements given for the smallest globules are only approximations, as they are really too small for measurement. Note that Koenig states, (Chem. Mensch. Nahr. Genussm., 4th ed. 2: 588) that the fat globules of cow’s milk range from 0.2 to 10 microns in diameter, average being 2 to 3 microns. RECOMMENDATIONS. No further modifications of the proposed new method are suggested. It is believed that the method promises very favorably to be a valuable one, especially in view of the growing tendency to incorporate foreign fats into creams and ice creams. The results are doubtless sufficiently uniform to warrant a positive recommendation that the method be adopted as provisional at this meeting. It is also recommended that the general continuous extraction method as described in this report and the modifi- cations for milk and cream products be given further study, especial attention to be given to the modification for rich cream and ice cream. SUPPLEMENTAL REPORT. By C. H. BIiesTeRFeLp. In the work on creams by the Paul method, the mass after extraction for 16 hours was reground and re-extracted for 5 hours, yielding the amounts of fat shown below, which were in each case less than 0.02 per cent on the charge taken. These amounts were included in the results previously reported. 1915] WHITE: CEREAL PRODUCTS 195 SAMPLE 2 SAMPLE 4 METHOD Weight Fat Weight Fat grams gram per cent grams gram per cent Paul’s method 1......... 27 0.0028 | 0.010 Aad | aise etal ltectiesgess Paul's method 2......... 50 0.0095 | 0.019 50 0.0081 0.016 W. F. Hand gave a short talk on a piece of special apparatus used in the analysis of dairy products. A paper by H. J. Wichmann on Lime as a Neutralizer in Dairy Products was read by G. E. Patrick. REPORT ON CEREAL PRODUCTS. By H. L. Wut, Associate Referee.' At a meeting of Committee C of this association held in 1912, the following recommendations, relative to methods for the analysis of wheat flour, weve adopted: (1) That the method of Bryan, Given and Straughn (Bur. Chem. Cir. 71) for soluble carbohydrates be given a further trial. (2) That methods for the estimation of moisture by the use of the vacuum oven and vacuum desiccator, for estimating the acidity of the water extract of flour and Olson’s method for dry gluten be referred to the next referee for immediate consideration. SOLUBLE CARBOHYDRATES. In following out these instructions two samples of flour were sent to each of six chemists who had offered to collaborate in the work. Sample 1592 was a hard red spring wheat patent. Sample 1244 was a straight flour made from a mixture of hard red spring wheat containing five per cent of sprouted wheat. Both samples gave good baking results, No. 1592 having a loaf volume of 2340 cc. and No. 1244 a loaf volume of 2345 cc. METHODS AND RESULTS. Soluble carbohydrates as dextrose. The method of Bryan, Given, and Straughn (Bur. Chem. Cir. 71), using both sodium carbonate and alcohol for extraction gave the follow- ing results: 1 Read by E. F. Ladd. 196 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 Comparative results on soluble carbohydrates as deztrose. SAMPLE 1244 : SAMPLE 1592 SSIES EAE . Extraction : Extraction Extract F F Extract = = with aleohol | Withsodium | with alcohol | With sodium per cent per cent per cent per cent R. F. Beard, Agricultural Col- lepescIN ED) eek sve oceetnetene its 11.49 1] 42 71.04 11.56 H. O. Lichtenwalter, Kansas 31.65 { 31-12 @ity3) Moe. sinn arceoeiscee tere 31.58 acest ay at 31.20 seen SM aa L. H. Bailey, Washington, D.C. 11.49 11.99 112 11.61 1 Average of 3 determinations. * Average of 4 determinations. * Results cal- culated from results expressed as invert sugar in next table. In connection with the work on carbohydrates, Mr. Lichtenwalter sent in the following results: EXTRACTION WITH ALCOHOL EXTRACTION WITH SODIUM CARBONATE SAMPLE no. |Reducing sugar} Total sugar Reducing sugar| Total sugar calculated as | calculated as Sucrose calculated as | calculated as Sucrose invert invert invert invert per cent per cent per cent per cent per cent per cent 1044 / 0.29 1.67 1.31 0.98 1.91 0.88 aes | 0.28 1.74 1.38 0.97 Lost Lost 1592 0.23 1.19 0.91 0.92 1.56 0.61 a 0.25 IL ePAS 0.96 0.89 1.60 0.67 Method I.—Use official method (Bulletin 107, Revised, page 38, 1), as a check. Method IT.—Vacuum desiccator: In an eighth-inch Hempel desiccator put 1 liter of sulphuric acid (specific gravity 1.84). Weigh out 9 samples of flour of 2 grams each into tared flat aluminum dishes, fitted with covers. Exhaust the air, using a vacuum gauge to indicate the pressure. (Some workers maintain a vacuum of 1 mm.; see J. Soc. Chem. Ind., 1913, 32: 72). In other determinations a vacuum of 15 mm. was maintained (Ibid., p. 69). Thoroughly mix acid and water layer twice each day. Remove and weigh 3 samples at the end of 3 days, 3 samples at the end of 5 days,and 3 at the end of 7days. Increase the time until constant weight is obtained. Method III.—Vacuum oven: Using 2 gram samples of flour, make a comparison of results obtained at 70°C. and 100°C. for 5 hours, maintaining the same vacuum, preferably 60 mm. Comparative results obtained by use of the different methods of deter- mining moisture are given in the following table: 197 CEREAL PRODUCTS WHITE 1916| ‘aanssoid Ammoieur “uu OT Ajazeutxorddy ; ‘ainsseid Aino1eul “Ulu ¢g Ajoyeutxoiddy iereele) Hf lusinsaiels sere Peers Peers lggerp | goa fccccccceteest emery copa | fpabeon) | Mocgoomel WercanE peer jo ever joo gor | ter | eet | lett | tocete | tere fo sess oN ‘ORR] -[0D [BingNosy ‘prveg ‘Wy “T €L°OT | 98°0T | 29°OT | 98°0T | “**** GS SeTT | ecseee| }OO nba 9 Ola | eae Sia Eas | ici rnmeea Ese Lama GOCE: Eom | OS aici T ct i OCT: a | cere eee ieee eo) Cat ‘uoWurysem ‘Aoeg “YT uaa dad |yuao wad |\quao sad |quao tad |yuao sad |\quao sad \uao sad \quao sad |quao wad \juao wad \quao tad |quao 49d |zuao sad |quao sad \auao sad quao tad sinoy g |Sinoy ¢ |sinoy g |Sinoy ¢| sup 6 | SAup J | SABp g | SABp g | SAUP gE | SABp 6G] SABp Z | SABP Q | SAUP g | SAuP g a]aureg) a a! Z6gT e[duEg FRET o[dueg 26ST e[dureg FRAT opduug Moats fa ISK1VNV Ill GOHLaN Il dOHLIN I dOoHLaN a aunjsiom UO sy)NsaL IayDIDdWOD 198 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 L. H. Bailey reports that on heating Sample 1244 one hour longer than in Method I, a loss of 12.07 per cent resulted; Sample 1592 under the same conditions showed a loss of 11.90 per cent. In Method II he main- tained a pressure of not more than 10 mm. of mercury with Sample 1592, and with some variation in Sample 1244. R. F. Beard began with an initial pressure of 3 mm. of mercury and re-exhausted desiccator on each day that samples were removed for weighing. At the end of the 9-day period, the pressure in the desiccator was 114 mm. of mercury. The results seem to indicate that 5 or 6 days in a vacuum desiccator under low pressure gives the maximum loss of weight. Method III was tried in vacuum oven at 70°C. and at a pressure of approximately 80 mm. of mercury. The results obtained at 70°C. are considerably lower than those obtained by other methods. GLUTEN. Olson method.—Dough up 10 grams of flour with 6 ce. of water. After weighing the wet gluten, place it in a vacuum oven and dry for 3 hours under 65 cm. vacuum at 85°C. The reduced pressure and the temperature combine to cause the gluten to expand rapidly. Compare the results with those obtained by the method given in Bureau of Chemistry Bulletin 122, page 54. Comparative results on gluten. SAMPLE 1244 SAMPLE 1592 Bulletin 122 Glzoutmethod Bulletin 122 ANALYST Olson method method method Moist | Dry Moist | Dry Moist | Dry Moist | Dry per cent| per cent| per cent| per cent| per cent| per cent) per cent) per cent R. F. Beard, Agricultural Colleges NEIDE ls neces: - se) oor oo | ARGS eer oe a(n 34256] 13-32) oval eee H. L. Wessling, Washington, IDR OREE Soe Omi aero MeO EOE aelliaocice LOLSi | eee. LOFS1 |) 2aeae 10.95) .....|/ 10:85 ACIDITY OF WATER EXTRACT. The acidity of the water extract of wheat flour depends on the time and temperature of extraction. Last year all collaborators agreed on 40°C. as being the best temperature to maintain, but disagreed on the time of extraction. It seems desirable to limit the time of extraction to two hours at the most; perhaps one hour will be more satisfactory. The method is as follows: Method.—Weigh out 18 grams of flour into a 500 ce. Erlenmeyer flask and add 200 ec. of distilled water free from carbon dioxid. Place the flask in a water oven kept at a temperature of 40°C. for the time indicated, shaking vigorously every half hour. Filter through dry double filters, rejecting the first 10 cc. of filtrate until 100 ce. 1915] MAGRUDER: VEGETABLES 199 is obtained. Titrate with twentieth-normal sodium hydroxid using carefully neutralized phenophthalein as an indicator. Each cubic centimeter of sodium hydroxid solution represents 0.05 per cent of acidity expressed as lactic acid. Comparative results on acidity of water extract. (Expressed as per cent of lactic acid). SAMPLE 1244 SAMPLE 1592 ANALYST 1 hour 2 hours 1 hour 2 hours isis Wee Wessling, Washington, TDA Ga eo as ba A 2 0.125 0.130 0.135 0.145 H. L. White, periculvural ¢ Col- lege, N. D... ; : 0.128 0.146 0.130 0.155 1 Used 2 ce. of phenolphthalein as indicator. RECOMMENDATIONS. It is recommended— (1) That the method of Bryan, Given and Straughn for the estimation of soluble carbohydrates be made official (results to be expressed as dextrose). (2) That the method for the estimation of acidity of water extract of flour given in this report, be made official. REPORT ON VEGETABLES. By E. W. Macrupemr, Associate Referee. CANNED FOODS. Since the associate referee for 1912 made no recommendations, it was decided to make some study of the percentage of easily separable liquid in some canned foods, such as tomatoes, corn, and butter beans. Time for this work, however, was limited and canned tomatoes was the only product studied. After some experimenting the regular fertilizer sieve with round holes 1 mm. in diameter was adopted as the best for separating the liquid from the solid portion. The method of procedure was as follows: Determine the weight of the contents of the can and then transfer the material to the sieve and stir the material gently with a spatula to allow the liquid to drain away; allow the material to remain on the sieve 5 minutes, gently stirring just before the expiration of the time. Then weigh the liquid and calculate the per- centage of separable liquid. At the end of the 5 minutes a considerable amount of liquid is still left with the tomatoes, but the great bulk of easily-separable liquid will have drained away. Five minutes seems to be about the right length of time to allow for drainage, as the object is not to get out all of the possible liquid, but that portion which is easily separable in order to determine whether water has been added and whether the tomatoes are of good quality. 200 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 Seventy-seven cans were examined in this manner; of that number 66 were the two, 6 the two and one-half and 5 the one and one-half pound cans. The average easily-separable fluid was 52 per cent, with the lowest 38 per cent and the highest 64.4 per cent. About 60 per cent came between 47 and 57 per cent of separable liquid. In the majority of cases each sample represented a different brand, but in many cases duplicate and triplicate samples were examined, and different cans of the same brand of the same pack were found to vary greatly. In one case the separable liquid varied from 44 to 69 per cent with a difference of 16 per cent, in another case from 47 to 57 per cent, while in others the variation was about 1 per cent. In all of the cans the tomatoes were in good condition. In some cases the tomatoes had been selected and prepared with more care than in others, but in no case was there found any evidence of added water, although it could not be told positively that none had been added. As a result of this work canned tomatoes which do not contain more than 50 per cent of separable fluid are considered by the associate referee to be excellent. Of course, tomatoes will vary from year to year in the amount of juice they contain and probably will vary at different seasons of the same year. A point which was not considered and which may have an influence on the separable liquid is the length of time the tomatoes have been canned. At the end of the work the total contents of the can and the separable liquid were measured in a graduated liter cylinder and the results agreed very closely with the results obtained by weighing. It is the opinion of the associate referee that the method of measuring could be substituted for weighing with advantage because of the saving of time which would thus be accomplished. This point, however, needs further study. It is reeommended— (1) That the associate referee for the ensuing year make determinations of the easily-separable liquid in canned tomatoes, corn, and butter beans, paying especial attention to the size and kind of sieve and to the time allowed for drainage, and in the case of tomatoes to the effect of the age of the canned product on the amount of easily separable fluid. REPORT ON COCOA AND COCOA PRODUCTS. By H. C. Lytueon, Associate Referee. MILK SOLIDS IN MILK CHOCOLATE. The present methods of the association for the determination of milk fat and lactose are satisfactory. The method of Baier and Neumann (Zts.. Nahr. Genussm., 1909, 18: 17) for the determination of casein was 1915] LYTHGOE: COCOA AND COCOA PRODUCTS 201 studied last year and proposed as a provisional method. Several samples of commercial milk chocolate were obtained and analyses made by my- self and my assistants, the results of which are as follows: Analyses of milk chocolate. 1 ny 2 a 2 |@n ne I B |Ba a SAMPLE ra] oz 4 ore a Key | IE! a NO. mi a Zo mo < a a ae Pees 6 E 8 iS ag ne i 8 $ wa lured) 2 Sore ee SB | eo Be eer Bo brese ea) a 3 a %, 3 & = x 3 B a 3 2 F Shae ee retievets 1.48 1.90 1.34] 3.14 3.92 | 33.09 | 4.51 5.64 | 38.63 | 4.84 5.70 | 0.69 | 15.26 ieceenatte 1.32] 2.30] 1.49] 3.90] 4.87] 32.04] 5.01] 6.18 | 36.08] 6.50] 7.04] 0.79 | 18.57 stefeiciar tel 2.06 | 2.26 1.92 3.69 | 4.62 | 24.12] 5.17] 5.82 | 44.89 6.22 | 6.72] 0.79 | 17.63 Mea eaeetele 0.70 1.97 1.26} 2.46 3.08 | 31.18 | 2.79 3.26 | 43.64] 4.35] 4.48] 0.94 | 11.37 Beyaiaararyarc 1.30 1.84 1.20] 3.29 4.11 | 33.93 | 4.29] 5.47 | 41.02 6.58 | 5.97] 0.75 | 17.20 aan aSRene 1.22} 2.00] 1.36] 4.00] 5.00] 30.16] 5.83] 6.84] 42.19] 8.47] 7.27] 0.73 | 21.62 Composition of milk used. a SOLIDS PROTEIN SAMPLE NO. SOLIDS FAT PROTEINS ASH NOT FAT LACTOSE FAT RATIO 15.03 5.59 3.89 0.75 9.44 4.80 0.69 13.70 4.56 3.59 0.75 9.14 4.80 0.79 13.59 4.48 3.56 0.75 9.11 4.80 0.79 12.55 3.60 3.40 0.75 8.95 4.80 0.94 12.54 3.99 3.00 0.75 8.55 4.80 0.75 12.26 3.88 2.83 0.75 8.38 4.80 0.73 CALCULATION OF THE COMPOSITION OF THE MILK USED. Milk proteins were calculated by multiplying the per cent of casein by 1.25. About 80 per cent of the total milk proteins is casein, the balance albumin. Milk sugar and ash are the least variable of all the constituents of milk. The sugar will vary in herd milk from 4.3 to 5.3 per cent, and from 4 to 5.8 per cent in milk from individual cows. The average milk sugar of 437 samples of milk of known purity examined during the past six years by the Massachusetts State Board of Health is 4.78 per cent, and of 47 samples of herd milk is 4.83 per cent. For the calculation of the composition of the original milk, 4.8 per cent was assumed to be the sugar content and 0.75 per cent the ash content and the computation was made by proportion. Thus in Sample 2, the milk sugar found was 6.50 per cent, the milk proteins 4.87 6.50 : 4.80 = 4.87:x x=3.59, the per cent of proteins in the milk. The sum of the fat, proteins, lactose, and ash gives the total solids and by proportion the total milk solids in the chocolate was obtained. For example, in Sample 2, the milk solids of the milk used was found to be 202 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 13.70 per cent, the lactose (assumed) 4.80 per cent, the lactose in the chocolate was 6.50 per cent. 4.80: 6.50 = 13.70: x x = 18.59 per cent of milk solids in chocolate. DISCUSSION OF RESULTS. The best comparison of the figures are those comparing the calculated values with actual milk analyses, which are shown in the following table: Analyses of milk. SOLIDS FAT PROTEINS ASH LACTOSE manana Samplettacaseter eee 15.03 5.59 3.89 0.75 4.80 0.69 Average of 10 samples ofmillkeesea- seer 15.00 5.62 3.75 0.76 4.87 0.67 Samplei2e. cscs sees eaellorn) 4.56 3.59 0.75 4.80 0.79 Average of 10 samples..| 13.70 4.59 3.40 0.75 4.77 0.74 Samplesteecce eee 13.59 4.48 3.56 0.75 4.80 0.79 Average of 10samples..| 13.60 4.67 3.40 0.75 4.78 0.73 Samples eecirerer eee 12.55 3.60 3.40 0.75 4.80 0.94 Average of 10 samples..}| 12.50 3.99 2.84 0.73 4.94 0.71 SEMM NG) Gosdaccevoosuacce 12.54 3.99 3.00 0.75 4.80 0.75 Average of 10samples..| 12.50 3.99 2.84 0.75 4.94 0.71 SamplelGseererecee ener 12.26 3.88 2.83 0.75 4.80 0.73 Average of 10 samples..| 12.30 3.89 2.81 0.72 4.88 0.70 According to the above comparisons all the calculated values are normal for milk of the same solids except 4, and the conditions found here may be due to the use of some skimmed milk. Assuming the average milk protein to be 3.30 per cent and the average sugar to be 4.80, the percentage of milk proteins multiplied by 1.454 would give the percentage of lactose. This ratio, however, varies with the composition of the milk used. The value with milk of 11.5 per cent of solids would be 1.8 and with milk of 14 per cent of solids would be 1.42. Therefore, not very reliable comparisons between the found and calculated lactose can be expected unless the milk used contained about 13 per cent of solids. The protein-fat ratios show that the methods for both casein and milk fat give results which are to be expected in products made from milk and are, therefore, reliable methods. The methods for casein and milk fat were tried upon a sample of malted milk with the following results: Fat 7.09 per cent, Reichert-Meissl number 27.6, casein 6.65 per cent. The method for casein applied to a proprietary food containing casein gave the following results: Total proteins (N X 6.38), 79.66 per cent; casein 79.68 per cent. It is recommended that the methods for the determination of casein and milk fat as proposed last year be adopted as provisional. 1915] BARTLETT: TEA AND COFFEE 203 REPORT ON TEA AND COFFEE. By J. M. Barrier, Associate Referee. As no report on methods of analyses for tea and coffee was given in 1912 the associate referee for 1913 was referred to the report of 1911, when a study of methods for the determination of caffein was made with recommendations that these methods be given further study. Therefore, the work has been confined to that subject. The Gorter and the modified Stahlschmidt methods, together with one suggested by H. C. Fuller of the Institute of Industrial Research, were selected as the most promis- ing for the work. Five chemists signified a desire to take part in the work and in May a sample of tea and one of coffee were sent them. These samples were prepared by grinding and mixing thoroughly several pounds of material which had been purchased on the market. The following instructions were sent with each set of samples: INSTRUCTIONS TO COLLABORATORS. Determine caffein or thein in each of the samples by the following methods: FULLER METHOD FOR TEA AND COFFEE. Weigh carefully 10 grams of No. 60 powder into a 500 ce. Erlenmeyer, add 100 ee. of water, 10 ec. of 10 per cent hydrochloric acid, and heat to boiling with a reflux condenser for 2 hours. Cool, decant liquid through a filter, treat solid material with 3 portions, 50 ce. each, of boiling water, filtering through same paper as just used and wash material on filter with 50 ce. of boiling water. Concentrate to 150 cc. by evaporating over steam or water bath. Transfer filtrate to a 500 cc. separator, Squibb type, add 5 cc. of stronger ammonium hydroxid, shake out with 50 ce. portions of chloroform five times. After the first shaking out let the separator rest until the separation is as complete as it will be; then run chloroform into another 500 ce. separator; add the second portion of chloroform and shake again, after standing until no further separation occurs, run the solvent and the adhering emulsion, if any, into the second separator, but do not run any of the nonemulsified liquid. Repeat three times, running chloroform and any emulsion into the second separa- tor. Then discard the liquid in the first separator and give the second separator containing the chloroform and emulsion a violent shaking; let stand and then run chloroform into a 250 ce. Erlenmeyer flask. If there is an emulsion remaining in the separator, add 1 to 5 cc. of 94 per cent alcohol and shake. When the chloroform has separated, add it to that in the 250 ce. flask. Add about 25 ec. of chloroform to the aqueous alcoholic layer in the separator and agitate; after separation run the chloroform into the 250 cc. flask and then evaporate off the chloroform on the steam bath using a moderate blast of air and removing from bath as last portions evaporate to avoid spattering. When the residue of crude caffein is dry add 10 ce. of dilute hydrochloric acid (10 per cent) and 50 ce. of water and warm until caffein is dissolved. Cool and precipitate with 50 cc. of iodin solution (10 grams of iodin, 20 grams of potassium oxid, 100 cc. of water), stopper flask with cork and let stand overnight. Filter through 9 cm. filter and refilter filtrate, if necessary, washing flask and pre- cipitate twice with iodin solution, but not attempting to remove all of the precipi- 204 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 tate from the flask. Transfer filtersto flask in which precipitation was made, add 0.5 gram of sodium acid sulphite, or sodium sulphite, 3 cc. of 10 per cent sulphuric acid and 15 ce. of water and warm until iodid is decomposed, more salt being added if the amount is insufficient to decolorize. Filter into a separator (small 100 ec.), add excess of stronger ammonium hydroxid and shake out 5 times with 15 ce. portions of chloroform. Wash the combined chloro- form extracts with water which is discarded, and then concentrate to 10 to 15 ec.; add dry animal charcoal, shake and allow to stand one hour with occasional shaking; filter through a small filter into a tared dish, wash flask and filter 3 or 4 times with 5 ee. portions of chloroform. Evaporate chloroform, dry residue in a desiccator and weigh. GORTER METHOD FOR COFFEE. Moisten 11 grams of finely-powdered coffee with 3 cc. of water, allow to stand for half an hour, and extract for 3 hours in a Soxhlet extractor with chloroform. Evaporate the extract, treat residue of fat and caffein with hot water, filter through a cotton plug and moistened filter paper, and wash with hot water. Make up the filtrate and washings to 55 ec., pipette off 50 cc. and extract four times with chloro- form. Evaporate this chloroform extract in atared flask and dry the caffein at 100°C. and weigh. Transfer residue to Kjeldahl flask with a small amount of hot water and determine nitrogen by Kjeldahl or Gunning method. N X 3.464 = caffein. MODIFICATION OF THE STAHLSCHMIDT METHOD FOR TEA. Boil 6 grams of finely-powdered tea in a flask with several successive portions of water for 10 minutes each, and make up the combined aqueous extracts thus ob- tained to about 550 ec. with water. Add 4 grams of powdered lead acetate to the decoction, boil for 10 minutes, using a reflux condenser, then add water so that the solution will finally be exactly 600 cc. Then pour the solution upon a dry filter and evaporate 500 cc. of the filtrate, corresponding to 5 grams of the tea, to about 50 ec. and add enough sodium phosphate to precipitate the remaining lead. Filter the solution, and thoroughly wash the precipitate, evaporate the filtrate and washings to about 40 ec. Finally, extract the solution thus concentrated with chloroform in a separatory funnel at least four times and evaporate the chloroform extract to dryness, leaving the caffein, which is dried to constant weight at 75°C. and weighed. The associate referee will be very glad if other methods are being used in the work of the analysts to determine caffein, to have results by such methods reported, and the methods given in detail. RESULTS OF COOPERATIVE WORK. COMMENTS OF ANALYSTS. H.C. Fuller: I would suggest that in running the Gorter method, when the aque- ous liquid is to be shaken out with chloroform it should be made ammoniacal in order to hold back certain coloring matter and facilitate in the extraction of caffein. In the Stahlschmidt method for tea, it seems to me that the directions are not suffi- ciently explicit when it states that the solution should be made up to exactly 600 ec. According to the directions given, this could be done either in the hot or cold solu- tion, and if it is done hot and it is attempted to filter 500 cc., the hot solution will have a chance to evaporate somewhat, and, of course, it is cooling all the time it 1915] BARTLETT: TEA AND COFFEE 205 is filtering—hence, becoming more concentrated, so that when you finally determine your caffein you will be working with a proportion which represents really more than 5 grams of tea. I think this is one reason why we obtained higher results by this method than by my own. Results of analysis of coffee and tea. CAFFEIN IN COFFEE CAFFEIN OR THEIN IN TEA db) ANALYST Gorter method ; Modifi euler Shablocheaae zptey Gravi- Nitrogen cu method GUL metric x 3.464 per cent per cent per cent per cent per cent H. C. Fuller, Institute of In- dustrial Research, Washing- TOTS ID So Oeee tesa tales velevevsiersiores. 1.270 10.850 1.170 3.140 2.850 3.306 | ~ 3.170 2.704 ay llleeece 3.134 | { 2.482 Frere 1.123 3178 F. F. Exner, Bureau of Chemis- 1.335 1.282 1.246 | | ————_ try, Washington, D.C...... 1.384 1.246 1.242 |) NX 3.4643 1.344 | | 1.220 3.073 2 3.054 1.158 Tae (3.054 H. B. St. John, Bureau of Chemistry, Washington, D.C.| #1.777 1.264 1.195 2.916 2.720 E. E. Sawyer, Agricultural Ex- 1.230 1.205 1.204 2.782 2.783 periment Station, Orono, Me. 1.220 1.23 1.213 2.800 2.819 IAM ETA PCY sere soho hen Latloeacs Severs 1.287 1201 eazy ll 3.055 2.735 1 Omitted from average. 2 Purified by shaking out with sodium carbonate solution. - B. H. St. John: It seems to me that the Gorter method could be successfully modified as follows: After shaking out the aqueous solution with chloroform, run the chloroform into a second separator and shake with a strong solution of sodium carbonate. The sodium carbonate solution will remove most of the coloring matter. Then pass the chloroform to a third separator and wash with water. Treat the re- maining chloroform shakeouts from the aqueous solution in the same manner, pass- ing them successively through the sodium carbonate and wash water. Unite the washed chloroform extracts, evaporate, dry and weigh, and put through the method of purification given in the Fuller method. I have given the results of one sample runinthismanner. The results obtained by the modification suggested, represents only a limited amount of work. From the analyses of the samples in question, however, together with the results obtained upon a number of samples of medicated soft drinks, the modification suggests itself to me as being worth further work to determine whether or not it will be of value. F. F. Exner: The Fuller method is laborious, time consuming, and by no means accurate. With tea the initial extraction with water is incomplete. This was proved by using the tea residue from two extractions in one portion, pouring more water over it, letting this stand on the steam bath for 2 hours, filtering and treating the filtrate as above for caffein; 0.028 gram of caffein was obtained. This would be 0.014 gram for each portion, which probably accounts for the low results obtained 206 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 with the tea. The first extraction with chloroform is difficult because of persistent emulsions especially with the tea. The caffein obtained from the tea is fairly white and pure, both by one sublimation, or by the iodin precipitation, so that the use of charcoal is not necessary. With coffee the first extraction with water seemed to be complete. Further extraction of the residue gave only a negligible amount of the additional caffein. The emulsion of the first chloroform extraction is not so persistent as is the case with the tea, but the caffein was still colored yellowish when purified by two sublime*ions or by the iodin purification. The two methods of purification are about equally efficient, but the sublimation is much shorter and simpler. The attempt to purify the caffein further by means of bone black proved unsuccessful. Though some of the color was removed, the caffein was not entirely white after this operation. It is very difficult to recover all the caffein upon filter- ing; although the charcoal and filter were thoroughly washed with chloroform, the 0.1246 gram of caffein from Experiment 2 with coffee, yielded only 0.1020 gram after treating with charcoal. In the Gorter method for coffee the initial extraction is convenient and complete. I used a large extractor with a large paper thimble 1} inches in diameter. I found it much better to remove the fat from the crude extract with petroleum ether, be- fore dissolving the caffein in hot water. This was done as follows: Pour 5 ee. of petroleum ether on the extract, dissolve the fat and decant the liquid through a small cotton plug in the stem of a small funnel. Repeat three times with 8 ce. portions of fresh petroleum ether; catch the filtrate in a small beaker and evaporate the ether off on the steam bath. Exhaust the caffein residue with boiling water according to the instructions, and filter through the same funnel and cotton plug, the filtrate being received in the separating funnel. Finally pour 5 cc. of hot water on the oil in the small beaker, heat the mixture and pour the whole onto the same filter at once. In this way all the water extract will run through the filter before it becomes clogged with the oil. A clean filtrate is obtained in which the subsequent extraction with chloroform is very rapid. Caffein is nearly insoluble in petroleum ether, but any traces that may have dissolved with the fat will be recovered by the subsequent treatment of the fat with water. There was a question in my mind as to whether the large amount of impurities in the residue from the second chloroform extract might not contain nitro- gen and so yield high results by the Kjeldahl method. I thought, therefore, to try a new form of sublimator which the writer has used with satisfaction on benzoic acid. It was thought that purification might be brought about and the caffein weighed directly. But the sublimate from the first sublimation was always of a pale sulphur yellow color, and after the second sublimation, a pale cream color. The second sublimate seems fully as pure as the results obtained by the Fuller method in the iodin treatment, but compared with the nitrogen estimation, the results are still high. Experiments 8 and 9 (see table following) show that the nitrogen determination is probably reliable, even when the crude, unpurified caffein is treated by the Kjel- dahl method, so the attempted purification by sublimation is not necessary. Indeed even though the results from the second sublimation are still somewhat high, there has been sustained a slight loss of caffein, for when the second sublimate is treated by the Kjeldahl method, the results are somewhat lower than in Experiments 8 and 9. Experiments 12 and 13 were for the purpose of checking the nitrogen method on pure material, and the results are entirely favorable. The modified Stahlschmidt method for tea proved on the whole quite satisfactory. T made the first extractions with 150 cc. of water and repeated the operation five 1915] BARTLETT: TEA AND COFFEE 207 times with 100 ce. of water. The lead acetate was added as directed before the last extraction was made. This last portion was used for the final filling to the mark. In this way the extraction was practically complete. To test this, the residues from the four experiments were combined and heated for several hours with water. The liquid was treated separately for caffein and only 4mg. were obtained. This would be 1 mg. for each portion. The caffein as first weighed had considerable color, and the result is doubtless high. One sublimation gave a product of a good white color, though the solution of the sublimafe in chloroform still had a slight yellow tint. Instead of the sublimation the nitrogen determination may also be used, and I believe we are justified in placing confidence in the result. The sublimation weight will doubtless always be slightly higher, but either method of purification may be considered satisfactory here. Results of sublimation of caffein in coffee. CAFFEIN IN CAFFEIN IN CAFFEIN PURI- METHOD | corrFEE| CRUDE NITRO- N X 3.4643 exrenient| USED |CAFFEIN Sea ron peetamiien Spanos EN SINDEN ee. grams | gram gram | percent) gram | percent| gram | percent] gram gram | per cent 8 Waescnd 0.1316 | 1.316 | 0.1288 Mert Fseeeaoll » stone cooseeee | eeuddad knagucic if) {| Se scnn: | iceadcor ieee | Peaocosall l moneae OF246 ie) 246)) | eestor rere ctor ieiate | rerereinia NY) fh ccoocos!|) openbe ||) Saree lp soasdn |) | séaee 0.1246 PLL | osondde|| cocondall Scaqos GORTER METHOD: Ms oecaseas 11 0.1931 | 0.1578 1.438 | 0.1468 DCR imeeane! |P emeere || ae doced| ltooonees| ll oacssao BE th aitet ll OP206 7a Os Lb8 701/90? 443!P Toss> [ee Yea || ractioniul]) oe noutl| sersisyesteclll, omeneocnoceeen Re Bas serie 1l 0.1540 | 0.1361 ORY dl Moorcdey || acted) peceaaell (hagacen Ss -abccd (tassoned ler haec ileousopacd ll LEGS Saa6oe ||apaobes ||. Geaone || “assaee NSdeneo |lki padge | (Sabedocd lyGucaned Ikdodadd (secgnaecse 11 0.1936 | 0.1539 1.399 | 0.1468 HAE ey eeoeocies| |b aseodo TOssin | Werner | recta f} Losoposes 1l 0.1852 | 0.1478 Neg? 7 Wa ts Bon Measiootc: || Booodd. | conabe 0.037614) 0.13031) 1.185 aaovecdase ll OF1614) |NOS1366) | eb 2424] Passe | tee cse |paaaircee ||) esis 0.035649) 0.12350) 1.123 yodacaans LR Saesared Moedoog ny leesae ed] Msacacnul (ence licesehe ciel | Mere ate 0.040702) 0.14100} 1.282 LW hempeaoae LGR | eectareterssta | asccie.= ston Ip catohetaeh |p catecetave.||fpmatsreveh |\tertnccse. |! wrelerte 0.039580) 0.13704] 1.246 OG enacts il 0.2330 | 0.1622 1.475 | 0.1522 URL epodes Ih conke 0.038740) 0.13420} 1.220 Ue aeedor ll 0.1994 | 0.1520 | 1.382 | 0.1478 MEREE S|! Gooede |) oodec 0.037333) 0.12733) 1.158 LP dee eae 0.3060 | c.p.anhydrouscaffeintaken | ......] ..... 0.087859] 0.30437) 99.460 ie babconepe, 0.2794 Clg I Wi Mossad | legasre 0.071000) 0.27714) 99.190 1 Loss in extraction; results not comparable, except to show relation between sublimation and nitrogen results. Experiments 1, 5, 8, 9, 10, 11, 12 and 13, should be given chief weight. No. 2 is allright as far as it goes. Results of sublimation of caffein in tea. Fuller method. CAFFEIN PURIFIED BY IODIN EXPERIMENT NO. TEA USED CAFFEIN IN FIRST SUBLIMATION PRECIPTUATION grams gram per cent gram per cent 1. OntintlereRRae 10 0.2542 QROAZ AMAT Soci aere tne |r aes eepee 2 obN50 BOO ORE SEIN LO ee erect sae IF Manet 0.2704 2.704 Tees sisre dé /ics, Fat iO Pl See eee | enact 0.2482 2.482 QIN area aie cle skeet 10 0.2724 DePQAP OND Lact A || ee a See 208 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 Modified Stahlschmidt method. CAFFEIN FROM 5 3 - = y ¥ FROM 3 v 2 waren | se | cave or wotourn | SPEED tno 3 | mims0OEN | 4x 3.4048 = carparn grams gram per cent gram per cent gram gram per cent Lopeasicn 6 0.1802 | 3.604 0.1653 | 3.306 | 0.044351) 0.15365 | 3.073 Dieta ase 6 0.1614 | 3.228 0.1585 | 3.170 | 0.044070) 0.15267 | 3.054 Gs datos 6 0.1594 | 3.188 0.1567 | 3.134 | 0.043228) 0.14976 | 2.995 Aye ee 6 0.1614 | 2.228 0.1589 3.178 | 0.044070} 0.15267 | 3.054 NOTE BY ASSOCIATE REFEREE. The results on the whole by all three methods are quite satisfactory with the coffee, but it is evident that the Fuller method gave rather low results on tea. It is evident that the Gorter method yields the eaffein in a somewhat impure condition and it is necessary to determine the nitrogen in the residue in most instances to arrive at a correct result. The Fuller method, on the other hand, yields the caffein in a fairly pure con- dition and in the ease of coffee the results obtained agree very closely with those obtained by the Gorter method when calculated from the nitrogen determined. If some modifications can be introduced into the Fuller method to facilitate the extractions and filtrations in the first part of the method, which are now very tedious and somewhat incomplete, particularly in the case of tea, it could possibly be made an accurate and desirable method for both tea and coffee. RECOMMENDATIONS. It is reeommended— (1) That the Fuller method for determination of caffein in tea and coffee be further studied with a view to improving the method of extrac- tion and filtration. (2) That the Gorter method be further studied with a view to purify- ing the caffein with sodium carbonate solution for direct weighing. (3) That the modified Stahlschmidt method be given another trial for the determination of thein in tea. THE SUBLIMATOR AND ITS USE. By F. F. EXner. The substance to be sublimed is placed in the bottom of the sublimation flask. It can be weighed in the flask, for this is composed of an ordinary beaker flask of light glass and 10 ounce capacity. The glass coil is fitted to slip easily through the perforations in the cork, yet the apparatus should not leak. The bottom of the cork which goes inside the flask is covered with a disk of tin foil. Through a third perforation in the cork is slipped a glass tube about 10 cm. long, its lower end being flush with 1915] EXNER: THE SUBLIMATOR 209 the bottom of the cork. The apparatus is supported on a wire gauze with asbestos, on a stand above a Bunsen burner. The ends of the glass coil are connected with rubber tubes to the tap so that a slow current of water will pass through the coil. The bottom of the flask is gently heated with a flame about 1 inch high under the asbestos. The caffein soon begins to sublime and condense on the cooled coil. When most of the substance has been driven from the bottom, a second flame, slightly yellow and rather large, but not too intense, is rapidly played around the sides of the flask until all has been completely volatilized and condensed on the coil. If some of the sublimate seems loose, the apparatus is shaken a little, and if anything falls off the coil it is again sublimed to the coil. When complete the apparatus is allowed to cool, the rubber tubes are disconnected, the apparatus carefully lifted from the stand, then the coil with the cork are lifted gently from the flask and placed in the Expansion Tube. — Cor k Weil Ain ; Tin Fovl. 8 i Tube. Sublimation Flask FIG. 2. a.-SUBLIMATOR. b.-WEIGHING TUBE. weighing tube. This is then supported on the stand and the stem of a funnel is attached to one coil and a double bend glass tube to the other. First alcohol, then ether, is run through the coil, and finally a current of air to dry it. The coil is then slipped from the cork, the tin foil being loosened and left with the coil. This is placed in a desiccator and weighed. Previous to the sublimation the apparatus was weighed in the same way. The difference gives the sublimate. The expansion tube. This has a small rather loose plug of cotton in the upper end, and is weighed separately before and after the sublimation. There is usually from 0.6 to 1.5 mg. of the sublimate in this tube. For benzoic acid which sublimes nicely from 120° to 140°C., the writer uses a glycerol bath. The caffein can be sublimed also in a glycerol bath, but the fumes are annoying. A convenient air bath could readily be devised, but the above procedure does so well that it is hardly worth while to use these accessories. 210 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMIsTs [Vol. J, No. 2 REPORT ON PRESERVATIVES. By A. F. Serxer, Associate Referee. The work this year has been confined entirely to the detection and determination of formic acid. An examination of the literature indi- cates that small amounts of formic acid may exist naturally in certain products or may be found in small amounts during the process of distil- lation required to separate it in the pure condition necessary for a qualita- tive test. It has been found! that formic acid is formed during the fer- mentation of sugars by yeast and is also decomposed during the same process so that little, if any, finally remains. The question of its presence in honey is still open, though the work of Fincke,? Farnsteiner,’? and Heiduschka and Kaufmann‘ shows that the amount present is in any event less than 0.02 per cent, many samples showing none at all. ‘It is also said to be produced by the slow oxidation of some terpenes in the presence of moisture® which would lead to traces being found in some prod- ucts flavored with essential oils. Further it has been shown that small amounts of formic acid are formed when carbohydrates are distilled in acid solution, Kreis expressing the opinion that qualitative tests are for this reason of no value and that quantitative results alone will indicate the presence or absence of added formic acid. In view of these state- ments contained in the literature, the plan adopted for this year’s work was as follows: (1) To select what appeared to be the most satisfactory method for determin- ing formic acid and to test its accuracy upon some simple mixtures. (2) To examine as many food products as possible with a view of ascertaining how much formic acid (or what appears to be formic acid) exists in them naturally or is produced in them by the operations required for its determination. (3) To submit to collaborators for analysis samples of food products containing known amounts of added formic acid together with a detailed description of the method for its determination. A study of the quantitative methods that have been proposed, a com- prehensive bibliography of which may be found in Fincke’s paper on “The Detection and Determination of Formic Acid’? would indicate that the reduction of mercuric chlorid and weighing of the resulting mercurous salt would yield the best results for practical purposes. The 1 Franzen, Chem. Ztg., 1911, 35: 1097; Franzen and Steppuhn, Zts. Physiol. Chem’ 1912, 77: 129. 2 Zts. Nahr. Genussm., 1912, 23: 255. 3 Thid., 1908, 15: 598. 4 Thid,, 1911, 21: 375. 5 Kingzett and Woodcock, Chem. News, 1912, 105: 26; J. Soc. Chem. Ind., 1910, 29: 791. 6 Fincke, Zits. Nahr. Genussm., 1911, 21: 13; Kreis, Mitt. Lebensm. Hyg., 1912, 3: 205; 266. 7 Biochem. Zts., 1918, 61: 253. 1916] SEEKER: PRESERVATIVES 211 method of Wegner! depending upon the formation of carbon monoxid by treating formates with concentrated sulphuric acid? has also been strongly recommended and may prove of value for confirmatory purposes, but appears to be too elaborate for general use. The procedure selected was that devised by Heinrich Fincke* which consists in distilling the material, acidified with tartaric acid, in a current of steam, passing the vapors through a boiling suspension of calcium carbonate in water, the vapors being then condensed and collected in order to measure the amount of vapor that has been passed through the sample. The calcium carbonate mixture is filtered and formic acid de- termined in the filtrate by reduction of mercuric chlorid. The points to be particularly observed in this procedure are: (1) That the volume of the sample shall not exceed 100 ce. and that it be kept at that volume throughout the distillation. (2) That not less than 1,000 ce. of distillate be collected. (8) That the substance shall contain no free mineral acid. (4) That sulphurous, sorbic, glyoxylic, levulinic, and fumaric acids are absent. Regarding the first two points Fincke found that with 100 ec. of sub- stance and 1 liter of distillate, 90 to 95 per cent of the formic acid is re- covered. By increasing the amount of substance or by decreasing the amount of distillate the percentage of recovery rapidly diminishes. These statements have been confirmed by the associate referee, as the following results show: Amount Formic acid of distillate. recovered. Substance cc. per cent. 0.0992 gram of formic acid, 100 cc. of water................... 200 52 0.0992 gram of formic acid, 100 cc. of water................... 500 85 0.0992 gram of formic acid, 100 cc. of water................... 1000 96 0.0992 gram of formic acid, 50 cc. of water..................+- 1000 98 0.0992 gram of formic acid, 150 cc. of water..................-- 1000 85 It has been found as pointed out by Fincke that formic acid is produced upon distilling solutions of carbohydrates containing free mineral acid, and methods based upon this procedure, several of which have been proposed, are certain to give high results. 1Zts. Anal. Chem., 1903, 42: 427. 2 Ost and Klein, Chem. Ztg., 1908, 32: 815; Merl, Zts. Nahr Genussm., 1908, 16: 389; Rohrig, Ibid., 1910, 19: 1. 3 Zts. Nahr. Genussm., 1911, 21: 1; 1911, 22: 88; 1912, 23: 255; 1913, 25: 386; and Biochem. Zts., 1913, 61: 253. 212 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 The following experiments by the referee show this to be true: Grams of formic acid Composition of solution distilled found. 50 grams of sucrose, 2 grams of tartaric acid, 100 cc. of water.......... 0.0024 50 grams of sucrose, 2 cc. of 85 per cent phosphoric acid, 100 cc. of water... 0.0102 50 grams of sucrose, 1 ce. of 85 per cent phosphoric acid, 50 ce. of water.... 0.0639 50 grams of sucrose, 5 ce. of 85 per cent phosphoric acid, 50 ce. of water... . 0.2157 A commercial chocolate fountain sirup (total solids 70 per cent, sucrose 58 per cent) in a dilution of 50 grams to 250 cc., acidified with 5 to 10 cc. of 2Oiper cent; phosphoridiacidteeeryaes aaeee are eee aces 0.1021 pane phocalate sirup in same dilution acidified with 2 grams of tartaric £106 | en ee ce er ceria tere Ad choca bao mr RnR Bl Sed En Be at 0.010 Concerning the nature of interfering substances and their removal, only the effect of the common preservatives, sulphurous, benzoic and salicylic acids have been studied. In his early work Fincke sometimes used a barium carbonate suspension instead of calcium carbonate for absorbing the vapors of formie acid, but in his later work for unexplained reasons, only calcium carbonate was used. It is well known that barium sulphite is comparatively insoluble in hot water and it was thought prob- able that by substituting barium carbonate for calcium carbonate in Fincke’s procedure the interference of sulphurous acid could be eliminated. As a test of this theory a mixture of 100 cc. of water, 2 grams of barium carbonate, and 5 ee. of a saturated solution of sulphur dioxid in water were mixed in the cold and filtered. The filtrate was reduced in the regular way and yielded the equivalent of 0.0037 gram of formic acid as mercurous chlorid. A mixture of the same composition as the above was then prepared and boiled before filtering, the last operation being conducted while hot. The reduced mercury salt amounted to the equiv- alent of 0.0010 gram of formic acid. The next experiments were conducted in the manner of the regular Fineke determinations using a suspension of barium carbonate instead of caletum carbonate. Reduction equivalent to grams of roe Composition of solution formic acid. 75 ec. of lime juice, 25 ce. RaW TST cee Sie Sees cs etn Mer ede 0.0026 75 ce. of lime juice, 25 ce. of water, 100 mg. of sulphur dioxid.. See 0.0032 75 cc. of lime juice, 25 cc. of water, 100 mg. of sulphur dioxid.. 0.0041 100 cc. of cherry juice, 2 grams of tartaric acid, 0.2 ae salicylic acid. 0.0026 50 grams of strawberry sirup, 0.05 gram of benzoic acid, 2 grams of tartaric EY To Bris Drevia, Chis) ae Se etn ero nae oa dae A ean michoe rosa dboeS 0.0015 With these preliminary tests completed the method as follows was adopted for trial: FINCKE METHOD FOR FORMIC ACID. APPARATUS. A device for generating steam. A 300 ec. round bottom flask (hereafter designated A). 1915] SEEKER: PRESERVATIVES 213 A 500 ce. round bottom flask (hereafter designated B). A condenser and a receiver in which the distillate may be measured. When in operation the steam passes through the liquid in flask A and then into the next flask by way of a spray trap and connecting tube the end of which reaches almost to the bottom of flask B and is furnished with a special tip consisting of a bulb containing a number of small holes for the purpose of breaking the vapor into small bubbles. From the second flask the vapor passes into the condenser and from thence the distillate passes into the receiver which is graduated to show a content of 1 liter. PROCEDURE. For thin liquids like fruit juices use 50 cc. For heavy liquids and semi-solids like sirups and jams use 50 grams diluted with 50 cc. of water. Place the sample in flask A, add 1 gram of tartaric acid (samples like lime juice which are naturally very acid need not be acidified), heat to boiling and steam dis- till, passing the vapor through a boiling suspension of 2 grams of barium carbonate in 100 cc. of water contained in flask B. If the sample contains much acetic acid more barium carbonate must be used in order that an excess of 1 gram remains at the end of the operation. The vapor issuing from flask B is condensed and measured. Continue the distillation until 1 liter of distillate is collected, the volume of liquid in flasks A and B being maintained as nearly constant as possible by heating with small Bunsen flames. After 1 liter of distillate has been collected disconnect the apparatus and filter the contents of flask B while hot, washing the barium carbonate that remains on the paper with a little hot water. The filtrate and washings should now measure between 100 and 150 ce. If not, it should be boiled down to that vol- ume. Add 10 ce. of sodium acetate solution (50 grams of dry salt per 100 cc), 2 ce. of 10 per cent hydrochloric acid, and 25 ec. of mercuric chlorid reagent (100 grams of mercuric chlorid, 150 grams of sodium chlorid dissolved in sufficient water to make 1liter). Mix thoroughly and immerse the container in a boiling water bath or steam bath for 2 hours. Filter on a tared Gooch, wash thoroughly with cold water and finally with a little aleohol. Dry in a boiling water oven for half an hour, cool and weigh. The weight of mercurous chlorid obtained multiplied by the factor 0.0975 gives the weight of formic acid present. If the weight of mercurous chlorid exceeds 1.5 grams the determination must be repeated using more mercuric chlorid reagent or a smaller amount of sample. A blank test should be conducted with each new lot of reagents employed in the reduction, using 150 ec. of water, 1 cc. of 10 per cent barium chlorid solution, 2 ce. of 10 per cent hydrochloric acid, 10 ec. of 50 per cent sodium acetate solution, and 25 ce. of mercuric chlorid reagent. The weight of mercurous chlorid obtained in this blank test should be deducted from that obtained in the regular determination. RESULTS. A standard solution of formic acid to be used in testing the method was prepared and standardized both by titration and by reduction. By titration. By reduction. Grams per 100 cc. Grams per cc. 0.00993 0.00990 0.00994 0.00995 0.00991 0.00990 Average, 0 00992 214 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 In the reduction method a blank reduction test was conducted upon the reagents and deducted from the individual determinations. The method applied to 50 cc. of aqueous solution containing 0.0992 gram yielded 0.0979 gram of formic acid or 98.7 per cent recovery. Further tests of the method gave the following results: FORMIC ACID AMOUNT OF MATERIAL TAKEN ADDED FORMIC ACID RECOVERED gram gram per cent 50 cc. of lime juice, 20 cc. of water.......... 0 0.0018 50 ce. of lime juice, 20 ec. of standard formic acid); ic. TA Se ee Ae ea 0.1984 0.1935 96.6 50 cc. of lime juice, 2 cc. of standard formic BOG cost we Ae aoe ones mp techs aver me 0.0198 0.0201 101.0 42 cc. of lime juice, 7.6 cc. of standard formic BOLGS ecpteysen cree Ae av Cee arte 0.0754 0.0740 98 .2 50 grams of strawberry sirup, 10 ec. of stand- ard formic acid, 40 cc. of water........... 0.0992 0.0946 95.4 The following determinations were made upon a variety of products in the course of routine inspection and represent the usual blank tests by this method when added formic acid is absent. The blank test on reagents has been deducted in each case. Determination of formic acid from inspection samples. FORMIC ACID FOUND IN GRAMS PER 100 cc. OR DESCRIPTION OF SAMPLE IN PER CENT BY WEIGHT Ney: 37030)Honey-Cuban (bulk)is-- eee sseee eee eee 0.0091 INeSY-157023 Honey-Cubant (bulls) sees ee or eeieiae eet: 0.0041 NewS 36715 Eoney-Cubann(bulle) seem ers smi icirseie eras 0.0029 INSeYe ro qOZoueloney-Gubann(Dulkc) seared eects stant erreicter- 0.0130 Cherry juice pressed in laboratory............-...-..---+-:- 0.0032 Strawberry sirup (sucrose 50 per cent) from juice pressed AN] BD OLATOLY: Merce Sitesese ope aorta eee ee Oa see Siege sree ee pias 0.0009 N. Y. 37814 ‘‘Forest Pearl’? summer drink sirup (in bottles). 0.0018 Ne NGBYGGL IAG UES: (MD) sag ccusa oso gsponoeesaouMoeHoUS 0.0026 INenYerss(Olelimejarcesan) botiles) sscmemea steerer 0.0019 N. Y. 39910 Lime juice fortified (in bottles)................. 0.0014 LSE PR NRT BI RAH eae pos cokeconn Hoebensgbosasoesoduas 0.0081 Ts529784. Crabvappleijellysccwse citing seca eles is a ens seeke 0.0051 PSS! 9786iGrape jellyy. oe cs sate terees tee nsisiqoe iereieeree oer 0.0083 N. Y. 40010 Raspberry juice (bulk)................--.0.00 0.0099 ING is.39038;) Blackberry, juices (bwlk) hn) 115 eleite teieiier inte 0.0034 INepYaue 90374 C herrya}UI1cen (bulk) Hanmi selenite aa 0.0043 NYA 391 742Prunesjuice (bulk) sa. ceree eee ce eeteeeeree 0.0067 N. Y. 39767 Raspberry sirup (in bottles).................... 0.0040 During the year one sample was met with containing added formic acid, N. Y. 37809, a cherry sirup imported from Denmark in bottles, analysis 1916] SEEKER: PRESERVATIVES 215 showing 0.148 per cent. Another N. Y. 38558, a raspberry shrub, con- tained 0.043 per cent of formic acid which probably entered the com- pound through the use of commercial acetic acid in its preparation. It was found upon examining a coffee extract, I. 5S. 8885 E (solids 17.49 per cent) that there was a notable amount (0.063 per cent) of what ap- peared to be formic acid present. Upon further investigation it devel- oped that coffee extract prepared in the laboratory from pure materials also contained formic acid, a solid extract made in this way having 0.627 per cent and a liquid extract (100 cc. representing 100 grams of origina coffee) 0.061 per cent. Qualitative tests gave strong reactions for formic acid, but whether the entire reduction is due to this substance will be further investigated. The referee has reason to believe from some uncom- pleted work that a similar reduction may be obtained from some roasted cocoa and roasted malt preparations, but this will be more fully reported later. That the partial caramelization of sugar will cause the formation of this reducing substance seems indicated by the results of a formic acid determination in which, through faulty regulation, the substance (grape jelly) was allowed to boil down too low during the distillation, the amount of formic acid found being 0.028 per cent, whereas in a determination properly conducted the result was 0.008 per cent. With respect to qualitative tests that of Fenton and Sisson! as em- ployed by Fincke? has been found exceedingly delicate, 0.5 mg. being detected without difficulty. The procedure is as follows: Introduce 10 cc. of the neutral or slightly acid liquid into a test tube, add 0.5 gram of magnesium ribbon in the form of a compact bundle, place a glass rod in the test tube upon the magnesium bundle in such a manner as to keep the magnesium at the bottom of the tube. Cool in a bath of ice water and add 6 cc. of 25 per cent hydrochloric acid a few drops at a time during an interval of about 30 minutes. When effervescence ceases test portions of the filtered liquid for formaldehyde by Leach’s test. As applied to sirups and fruit juices the filtrate from the barium carbonate suspension as obtained in the quantitative determination, can be boiled down to 10 cc. for this test, or 100 to 150 cc. of an ordinary steam distillate of 50 to 100 grams of the substance are treated with barium carbonate, filtered, and the filtrate boiled down to 10 ce. The difficulty in employing this test is its delicacy, and owing to the fact that small amounts of formic acid may be present in some natural products or are formed during the process of distillation it can not be relied upon as a means of detecting added formic acid. The associate referee has obtained positive tests with some cocoa, coffee, wine, vinegar, and with sucrose distilled in an aqueous solution made slightly acid with phosphorie acid. 1 Proc. Cambridge Phil. Soc., 1908, 14: 385. 2 Biochem. Zts., 1913, 61: 259. 216 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. J, No. 2 The conversion of formates to hydrocyanic acid has been tried by means of ammonium sulphate and phosphoric anhydrid, but both it and the Grosse-Bohle reaction! which consists in treating 10 ce. of the solution, first with 1 to 2 ec. of 25 per cent hydrochloric acid and then with fuchsin sulphite (0.1 per cent fuchsin hydrochlorid) have not been found to give satisfactory results. COOPERATIVE WORK. The collaborators this year were asked to test the accuracy of the Fincke method for the determination of formic acid. With this object in view sets of three samples were sent to each of fourteen collaborators, and seven replies were received, those cooperating being: (1) H. C. Fuller, Washington, D. C.; (2) A. L. Knisely, Portland, Ore.; (3) E. R. Ly- man, Portland, Ore.; (4) F. L. Shannon, Lansing, Mich.; (5) A. W. Hanson, Kansas City, Mo.; (6) R. W. Hilts, Seattle, Wash., and (7) C. H. Robinson, Ottawa, Can. The referee wishes to express his acknowl- edgments to these collaborators for their assistance in this work. The composition of the three samples sent to the collaborators was as follows: Sample A: 302 ce. of standard formic acid solution made up to 2 liters with commercial lime juice. The lime juice employed (N. Y. 37761) was part of a ship- ment in bulk received at the port of New York, and previous to being used was clari- fied by mixing with powdered asbestos and filtering. Formic acid added, 0.1498 gram per 100 ce. A blank determination upon the original juice gave 0.0026 gram per 160 ce. Sample B: 302 cc. of standard formic acid and 1.5 grams of sodium salicylate made up to 2 liters with cherry juice. The cherry juice was pressed in the labora- tory from red cherries and clarified by mixing with powdered asbestos and filtering. Formic acid added, 0.1498 gram per 100 ce. A blank determination upon the original juice gave 0.0032 gram per 100 cc. Sample C: 353 ec. of standard formic acid solution and 60 ec. of alcohol made up to 2,544 grams with strawberry sirup. The strawberry sirup employed was pre- pared by mixing 1 liter of fresh strawberry juice, pressed in the laboratory, with 1,900 grams of granulated sugar, bringing to a boil, and straining through muslin. Formic acid added, 0.1377 per cent by weight. A blank determination upon the strawberry sirup alone gave 0.0012 per cent. 1 Zts. Nahr. Genussm., 1907, 14: 89. 1916] SEEKER: PRESERVATIVES 217 RESULTS. The results obtained by the collaborators are given in the following table: Cc6perative results on determination of formic acid. SAMPLE A SAMPLE B SAMPLE C ANA- LYST Formic acid recovery Formic acid recovery Formic acid recovery gram per cent gram per cent gram per cent Merete; ste 0.1454 97.1 0.1416 94.5 0.1260 91.5 PA Oe 0.1496 99.9 0.1506 100.5 0.1366 99). 2 Seria 0.1484 99.1 0.1468 98.0 0.1300 94.4 (a 0.1470 98.1 0.1480 98.8 0.14 101.6 0.1500 100.1 0.1500 100.1 0.13 94.4 Diaisvas 0.149 99.5 0.153 LODE oe ater ess rele 0.105 70.1 0.143 a ae iss orice aes Bioko ac: 0.153 102.1 0.151 100.8 0.135 98.0 0.153 102.1 0.150 101.1 0.136 98.7 Tereave! 0.142 94.8 0.1388 92.7 0.142 103.1 0.144 96.1 0.13874 91.7 0.139 100.9 DISCUSSION. R. W. Hilts in commenting upon the method says: I have used this method before, or an adaptation of it, in my work, and have also used the method of Wegner as described by Rohrig and found the two methods to check very wellindeed. The latter method is too elaborate for routine employment. Would it not be well to have a qualitative confirmatory test in addition to the quan- titative method? I have used the following, based on Bacon’s suggestion in Cir- cular 74: Distill about 25 ce. of the sample as in the quantitative method and boil down the filtrate from the suspended carbonate to about 30 cc. (or the simpler method may be used of receiving the distillate in a flask, adding the carbonate, and boiling down and filtering). Put in a small flask. Prove the absence of formaldehyde in this solution by testing a few cubic centimeters; add 1 gram of magnesium ribbon; attach an air condenser, stand in cold water and maintain an evolution of hydrogen for 1 hour by the gradual addition of hydrochloric acid, in small quantities. Test the solution for formaldehyde by the Leach or Hehner test. For qualitative pur- poses only about 300 ce. of distillate need be collected. Two or 3 mg. of formic acid can be detected after reduction in this manner, in my experience. Of course, a qualitative method such as the above should be well tested out on various fruit juices before being actually proposed for adoption. There have been no unfavorable comments upon the method from the collaborators. The results obtained by those who have coéperated are to be regarded as particularly gratifying and prove that the Fincke method is accurate for the determination of formic acid in ordinary fruit juices and sirups. The referee feels, however, that before it can be adopted as a general method a further study should be made of interfering substances. It 218 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 appears that roasted or partially caramelized substances give high results and it is recommended that a further study of the method with reference to these as well as the interfering substances mentioned by Fincke! be made during the coming year. It is also recommended that the collection of data regarding the amount of volatile reducing substances occurring in natural products as determined by this method be continued, and that nat- ural products be examined qualitatively for formic acid in order that the value of a qualitative test in detecting added formic acid may be finally ascertained. REPORT ON WATER IN FOODS. By W. J. McGesr, Associate Referee. The work for 1913 has been along the line recommended by the asso- ciate referee on this subject for 1911 and 1912, that is, the study of differ- ent dehydrating agents in drying food material in a vacuum. The inves- tigation was extended, however, to include the study of dehydrating agents at atmospheric pressure in direct comparison with the vacuum. The highest vacuum in this experiment was about 29 inches, with an aver- age of about 27, produced by a Schutte and Koerting water jet pump working at about a thirty-pound head. The dehydrating agents tried were: Sulphurie acid, phosphoric pentoxid, calcium carbid, metallic so- dium, calcium chlorid, barium peroxid, and glycerin. The foods experi- mented upon were: Cheese, cocoa, and cornmeal, in the first experiment and cheese, cocoa, and raw sugar in the second experiment. The following tabulated lists show the results of the experiments. A study of these results seems to show that sulphuric acid, phosphorus pentoxid, calcium carbid, and sodium metal are about equal in dehydrat- ing power. Therefore, the choice for laboratory purposes may be made on the basis of other properties. Sulphuric acid is not easily portable in a desiccator, and is perceptibly volatile when very hot dishes are placed over it, and also under other conditions which have been studied by H. C. Gore and others. Phosphorus pentoxid is portable, but in the desiccator an impervious layer seems to form on top, which retards its action. Calcium earbid is portable and the water absorbed converts the carbid into powder, which is easily shaken to the bottom of the receptacle. The acetylene gas formed should be given a vent with a mercury or other seal. It is possible, of course, that some classes of food would absorb acetylene gas or form additional products with it. This is not the case, however, with any of the foods so far tried. 1 Biochem. Zts., 1913, 51: 278. 1915] McGEE: WATER IN FOODS 219 Drying in desiccators over various dehydrating agents at room temperature. Experiment 1. SULPHURIC ACID / CALCIUM CARBID SUBSTANCE TIME Atmospheric pressure 76 mm. pres- sure Atmospheric pressure 76 mm. pres- sure days | per cent} per cent) per cent ms 1 25.50 2 2 26.34 aaa 3 26.49 Es 4 26.53 BS 5 26.44 Bs 7 | 126.56 dic 9 26.70 Z 11 26.51 a ee ae S al ee 1 3.44 x 2 5.63 7 3 6.00 Els 4 6.22 s 3 5 5.85 Bs 7 16.25 Pa 9 6.64 g 11 6.32 38 reo) ae (ant ena 1 7.54 5 2 10.66 ea 3 11.45 38 4 11.83 ate 5 | 11.58 ga 7 | 112.04 83 9 12.57 5 a 11 12.57 'é) re (ee (70-0 Bere METALLIC CALCIUM GLY- eas SODIUM CHLORID CERIN | oxip co 1 o 1 o °o i Hig z a2 ae | a2 a5 5 as si as as 82 | #2 | 82 | do | 823] 82 gi | £2 | 62 | £2 | s2 | 88 = Ree | Sails =< per cent) per cent) per cent| per cent| per cent| per cent| per cent 24.61 | 25.56 | 24.03] 25.85 14.93 25.34 | 25.64) 25.75 | 25.92 22.40 25.55 | 25.78 | 26.02 | 25.98 23.55 25.64 | 25.85 | 26.06 | 26.02 23.63 25.66 | 25.89 | 25.95 $ 23.97 25.84 2 CS Anan 24.18 25.90 : PPL | epee 24.22 25°97) |) 2497 265118 | Meecen |) rciceu|t eee 25.79) |) seca AUN CGrees ||) qoobe.|| ‘ado 26145 | eeeet Fi REM) ora || cacoe ||! Goose 2.05 4.75 1,32 4.65 0.14 0.21 3.52 5.33 3.96 4.81 1.06 0.46 4.12 5.58 4.67 4.83 1.71 0.80 4.75 5.76 4.86 4.90 1.95 0.91 4.52 5.71 4.90 3 2.03 0.94 4.91 2 5.14 2.27 1.07 5.05 2. 5.16 2.29 1.05 5.07 4.87 BN ee Gacel| nescil| soncos 4.86 UPRY || “Saace Il faccoe'|) ooode AUS “cansa De eauda ll) Ssecoan| codec 5.42 9.65 4.26 9.20 2.05 1.40 8.14] 10.61 8.52 9.50 3.74 1.82 9.21] 11.04 9.69 9.52 5.14 2.62 9.83 | 11.35} 10.12 9.62 5.62 2.95 9.98} 11.50] 10.24 2 5.79 3.07 10.68 2 LE Seger 6.11 3.37 11.03 2 LOSS Tale tees 6.26 3.36 LEAS) |) 105695 LOLS 7a ee eters | aresmioa | eteretey DD. 49) [0 eras 1h) eae Onan Maposan |, hemood 1 Acid renewed. 2 Vacuum off for 45 days. 4 Desiccator broken. Metallic sodium forms a scale of sodium hydroxid, which, however, There is also danger of too much rise in temperature when a substance having a high percentage of seems not to impair its dehydrating power. water is placed over it. jections and mechanical difficulties as is acetylene gas. acetylene gas and hydrogen are not usually serious, and for ordinary laboratory work it would seem that, all things considered, calcium carbid or sodium metal are the most practical reagents. The hydrogen formed is subject to the same ob- The objections to 220 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 Drying in desiccators over various dehydrating agents at room temperature. Experiment 2. SULPHURIC ACID CALCIUM CARBID Tasca SODIUM METAL SUBSTANCE | TIME phonic | 781m. | phene | 7mm. | Dhene | 7mm. | Dhoge | 76mm. pressure 3 pressure pressure | PT®SSUTC | pressure | Pressure days per cent | percent | percent | percent | percent | percent | percent | per cent 1 28.38 29.08 28.32 28.85 28.46 29.35 28.48 28.88 S a 29.06 29.35 28.87 28.94 28.87 29.40 28 .93 29.03 58 3 29.22 29.38 29.17 29.34 29.03 29.60 29.08 29.25 a 3s 4 29.30 29 54 29.22 29.40 29.06 29.33 29.13 29.32 £4 8 5 29.28 29.41 29.22 29.48 29 13 29.64 29.13 29.13 Fa Bg 5 6 29.30 29.42 29.25 29.44 29.16 29.64 29.15 29.38 ao a 8 29.46 29.38 29.41 29.56 29.26 29.66 29.22 29.45 ra 6 10 29.54 29.54 29.51 29.56 29.33 1 29.27 29.37 g 16 29.57 129.54 29.63 129.50 29.41 29.79 29.36 29.52 35 29.40 29.57 29.58 29.44 29.26 29.86 29.32 29.66 1 4.71 6.57 4.52 6.47 5.14 6.90 5.09 5.95 & 2 5.95 6.40 5.88 6.51 5.89 6.65 5.85 6.05 a 3 6.30 6.89 6.25 6.98 5.73 7.15 5.97 6.38 50 4 6.41 7.19 6.44 7.09 6.15 6.81 6.04 6.48 2 8 5 6.30 6.91 6.42 7.16 6.21 7.06 6.08 6.16 gs 6 6.42 7.06 6.51 (feilst 6.32 7.09 6.11 6.66 oa a 8 6.79 6.8 6.85 7.23 6.56 7.14 6.31 6.61 S 10 6.88 7.22 6.85 7.33 6.44 1 6.34 6.51 38 | orto 17.15 7.22 17.33 6.81 17.48 6.52 16.55 SOG. Tl Meyares-fe 7.28 7.44 7.39 6.74 7.36 6.58 6.95 2 1 0.96 0.87 0.85 0.87 0.96 0.94 0.94 0.86 5 a 2 0.99 0.85 0.94 0.90 0.99 0.98 0.99 0.92 gee 3 1.05 0.95 0.99 0.95 1.04 0.98 1.03 0.98 Ses 4 1.06 0.99 1.00 0.98 1.06 0.99 1.04 0.99 = 3 3 5 1.05 0.98 1.01 0.99 1.07 1.00 1.04 1.00 3 Bs 6 1.04 1.01 0.98 1.00 1.06 1.03 1.03 1.01 2am 8 1.08 1.03 1.01 1.01 1.07 1.01 1.06 1.02 E & 10 1.08 1.04 1.01 1.03 1.09 1 1.06 1.05 ps ° he NI Sea 10.99 1.02 11.00 1.07 11.01 1.07 11.01 Bir Milt odes 0.99 0.99 1.01 1.05 0.98 1.05 1.02 1 Vacuum off. The dehydration in a partial vacuum did not show in every case the advantage that was expected over the dehydration at atmospheric pres- sure. On both experiments in drying cheese, from 90 to 94 per cent of the water was dried out in 24 hours in the desiccators held at atmospheric pressure, while in the desiccators at a vacuum of 27 inches, the percentage ran from 93 to 97 per cent of the total moisture in the cheese. On the drying of raw sugar, the vacuum did not appear to be an advantage. Over sulphuric acid, phosphorus pentoxid, and sodium metal the highest results in 24 hours were from the desiccators at atmospheric pressure, while only in the calcium carbid desiccator were the results under vacuum better than those under atmospheric pressure. In drying cocoa and 1915) FORBES AND WUSSOW: INORGANIC PHOSPHORUS ESTIMATION 22] cornmeal, the vacuum method had a distinct advantage with all the re- agents tried. From the foregoing, it may be seen that dehydration in a partial vacuum is not always more perfect than over the same medium at atmospheric pressure. A fact that stands out clearly in the experiments is that in a desiccator at room temperature and with atmospheric pressure in 48 hours the ma- terials used for the tests lost from 80 to 96 per cent of all their water and at a pressure of 75 to 100 mm. they lost as much in 24 hours. The following recommendations are made, therefore, for future work: RECOMMENDATIONS. It is reeommended— (1) That comparison of drying organic or other materials at room temperature in partial vacuum and at atmospheric pressure, using phos- phorus pentoxid and calcium carbid as dehydrating agents, be continued. (2) That there be further comparison of the dehydrating power of sul- phuric acid, phosphorus pentoxid, calcium carbid, and metallic sodium, and any other reagent that may be found, at room temperature and atmospheric pressure. (3) That it be ascertained if a general method for moisture may not be used, consisting of 24 or 48 hours storage over either sulphuric acid, phos- phorus pentoxid, calcium carbid, or metallic sodium, at room temperature and atmospheric pressure to be followed by the vacuum oven at 70°C. or 100°C. for a short time. (4) That the moisture determination by vacuum method over sulphuric acid be made an optional official method. REPORT ON INORGANIC PHOSPHORUS ESTIMATION IN PLANT AND ANIMAL SUBSTANCES. By E. B. Forses, Associate Referee and A. F. D. Wussow. This investigation of methods of inorganic phosphorus estimation in- volved a comparison with animal substances of the neutral molybdate method of Emmett and Grindley, the barium chlorid method of Siegfried and Singewald, and the magnesia mixture method of Forbes and asso- ciates; and with plant substances of the acid alcohol method of Forbes and associates and the method of R. C. Collison, likewise founded on the acid-alcohol separation of phytin and inorganic phosphate. The work was carried through the years 1912 and 1913. Those co- operating with the authors of this report were in 1912 P. F. Trowbridge and A. C. Hogan of the University of Missouri, H. S. Grindley and E. L. Ross of the University of Hlinois and H. L. White and R. F. Beard of the 222 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMIsTs [Vol. I, No. 2 North Dakota Agricultural College; in 1913, H. 8. Grindley, C. I. Newlin, P. F. Trowbridge, O. C. Smith, and H. L. White. The general method of test of the several analytical procedures was by making the estimations with and without the addition of known amounts of inorganic phosphorus, the correctness of the methods being judged from the completeness of recovery of the added phosphate. In the work with plant substances attention was also given to completeness of extraction and the influence of phenol in the extractive reagent. The main body of this report is made up of the detailed specifications governing the work and the tabular presentation of results of analysis. The discussion of the results will be found on pages 236 to 239. DETERMINATIONS ON ANIMAL SUBSTANCES, 1912. Outline showing methods of precipitation, amounts of extracts to be used for each determination, number of determinations, and amount of standard phosphate solution to be added to certain of these determinations; the same set of determi- nations to be made on muscle, blood and brain: a-l 500 ce. of extract. { Neutral molybdate | a-2 500 cc. of extract. precipitation a-3 250 ec. of extract + 25 cc. of phos- (Emmett and phate solution. Grindley) a-4 250 ce. of extract + 25 cc. of phos- | phate solution. b-1 500 ce. of extract. Muscle Rarium chlorid b-2 500 ce. of extract. Blood precipitation b-3 250 ce. of extract + 25 ce. of phos- Brain (Siegfried and phate solution. (5,000 cc. extract) Singewald). b-4 250 ce. of extract + 25 cc. of phos- phate solution. pu 500 ce. of extract. -2 500 cc. of extract. 250 ec. of extract + 25 ee. of phos- phate solution. c-4 250 cc. of extract + 25 ce. of phos- phate solution. precipitation (Forbes and as- sociates). Magnesia mixture is | PREPARATION OF A COLD WATER EXTRACT OF DESICCATED FLESH. Weigh out about 45 grams of the vacuum-dried meat, and divide it among sixteen 150 ec. beakers; to each beaker with its contents add about 3 to 5 ee. of distilled water; break up any lumps and stir well with a glass rod until the mass forms a thick paste. Add 50 ce. of distilled water to each beaker and stir thoroughly for 15 minutes; allow the insoluble portion to settle for a few minutes (3 to 5) and decant the super- natant liquid through wet 11 em. filters; collect the filtrates in 250 ce. Florence flasks; take care that the funnels touch the sides of the necks of the flasks; drain the residues thoroughly, keeping as much of them as possible in the beakers; treat these residues with 25 ec. of distilled water, stirring for 5 to 7 minutes, and then allowing 9 1915| FORBES AND WUSSOW: INORGANIC PHOSPHORUS ESTIMATION 223 3 to 5 minutes for the solid particles to settle before filtering. Decant, etc., as just described. Repeat this last treatment until the filtrate measures about 220 cc., then transfer the entire residue to the filter and wash twice with about 8 to 10 ec. of distilled water. Allow all the liquid to pass through the filter before adding the next extract. Whenever the major portion of the residue has become mechanically transferred to the filter, return it to the beaker, using great care not to break the filter paper. Transfer the sixteen filtrates of about 250 cc. each to a measuring flask. Wash out each Florence flask twice, using about 5 to 8 cc. of distilled water each time. Make the extract up to 5,000 cc. and mix it thoroughly without too much mechanical agitation. PREPARATION OF HOT WATER AMMONIUM SULPHATE EXTRACT OF BLOOD. Weigh about 50 grams of fresh blood or its equivalent of oxalated blood into each of six 400 ce. beakers. To each beaker add a few cubic centimeters of distilled water and work up the blood and water with a glass rod; make up to about 200 ce. with boiling distilled water; place over a flame and gradually bring to boiling, with con- stant stirring. When boiling begins add to each beaker 20 cc. of 20 per cent ammon- ium sulphate solution; boil with constant stirring for about 10 minutes; decant onto sand on linen. When the liquid is through lift the coagulum off from the sand and return it to a mortar; grind to a smooth paste and transfer from mortar to beaker with boiling distilled water; make up to about 80 cc. with boiling distilled water; stir for 8 minutes and pour contents again onto the sand filter. After the extract is through, return the coagulum to the mortar and grind a second time, transferring to the beaker as before with boiling distilled water. Repeat this process of 8-minute extractions of the coagulum in hot water and filtration as just directed, without further grinding, until the filtrates measure about 750 cc. each. Wash out each beaker twice with 8 to 10 cc. of hot distilled water, completing the transfer of the coagulum and extract to the sand. Wash the coagulum on the sand twice with boil- ing water from a wash bottle. At all times allow the filter to drain well between additions of extract or wash water. Combine the six filtrates of about 800 cc. each, washing out the containers of each twice with distilled water. Make the extract up to 5,000 cc. and mix. PREPARATION OF HOT WATER AMMONIUM SULPHATE EXTRACT OF BRAIN. Weigh out about 10 gramsof brain into eachof ten 250 ce. beakers. Toeachbeaker add a few cubie centimeters of distilled water and work up the brain and water with a glass rod; make up to about 100 cc. with boiling water; place over a flame and gradually bring to boiling with constant stirring. After boiling has begun add to each beaker 20 ce. of 20 per cent ammonium sulphate solution; boil for about 10 minutes; allow to settle for a moment and decant liquid onto sand on linen. If the extracts do not filter readily, carefully push the coagulum to one side or return to the beakers. Add to the beakers containing the coagulum 50 cc. of 0.1 per cent ammonium sulphate solution; stir for 1 minute and decant the liquid onto the filter. Repeat this process of one-minute extractions of the coagulum in 0.1 per cent am- monium sulphate solution and filtration as just directed until the filtrates measure about 450 ce. Wash out each beaker twice with 8 to 10 cc. of hot 0.1 per cent ammon- ium sulphate solution, completing the transfer of the coagulum and extract to the sand. Wash the coagulum twice with the above wash solution from a wash bottle. At all times allow the filter to drain well between additions of extraet or wash solu- tion. Combine the ten filtrates, washing out the container of each of the filtrates twice with 5 to 8 ce. of distilled water. Make the extract up to 5,000 ec. and mix. 224 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 EMMETT AND GRINDLEY NEUTRAL AMMONIUM MOLYBDATE METHOD FOR INORGANIC PHOSPHORUS IN WATER EXTRACTS OF FLESH. Measure out the number and volumes of extracts specified in the schedule of determinations. Evaporate with frequent stirring on the water or steam bath to approximately 20 to 25 ee.; while hot, filter into 300 cc. beakers, using doubled 11 cm. No. 589 (Blue Ribbon brand) S. and S. papers. Wash beakers, precipitates, and filters thoroughly with hot water; the volume of the resulting filtrate and washings should be about 125 ec. Add 10 grams of ammonium nitrate and heat upon the water bath to 60°C.; add 10 ce. of nitric acid (specific gravity, 1.20), stir, and add 125 ce. of clear neutral ammonium molybdate solution. (Neutral ammonium molyb- date is prepared by adding ammonia to the ordinary molybdic solution, using lit- mus paper as an indicator. This work should be done very carefully and both red and blue litmus paper used.) Reheat, bringing temperature to 60°C. Keep at this temperature for 15 minutes, stirring vigorously every few minutes. Remove from the bath and allow the solutions to stand 2 hours in a warm place. Decant the clear supernatant liquid through doubled 11 em. No. 589 (Blue Ribbon brand) S. and §. filters. Transfer the remaining liquid and precipitate to the filters, using a 10 per cent ammonium nitrate solution. Wash precipitates and beakers four or five times with small volumes of the ammonium nitrate solution. Dissolve the yellow precipitate upon the filter and that in the precipitating beaker with dilute ammonium hydroxid (2.5 per cent) and hot water, collecting the filtrate in a 250 ce. beaker; wash thoroughly; neutralize the solution with nitric acid (specific gravity 1.20) and make up to approximately 150 ecc.; add 5 grams of ammonium nitrate; heat upon the water bath to 60°C. and then carefully add, while stirring, 5 cc. of concentrated nitric acid and 50 ce. of clear acid molybdie solution. Digest at 60°C. for 15 minutes, stirring occasionally. Continue the determination of phosphorus as usual weighing the phosphorus as magnesium pyrophosphate. SIEGFRIED AND SINGEWALD METHOD AS USED BY GRINDLEY AND ROSS FOR INORGANIC PHOSPHORUS IN ANIMAL SUBSTANCES. Measure out the number and volumes of extracts specified in the schedule of determinations. To each portion add 50 ce. of a 10 per cent barium chlorid solution and 10 ee. of a 10 per cent solution of ammonium hydroxid. Stir the solutions every 15 minutes for a period of 1 hour, allow to stand undisturbed for at least 12 hours, and then filter (decanting at first as much as possible) through double cuantitative filters. Wash the beakers, precipitates and filters, repeatedly, with small quanti- ties of wash water containing 10 cc. of the barium chlorid solution and 10 ce. of the ammonium hydroxid solution per liter. Place the upper filters containing the precipitates in the beakers in which the precipitation occurred, and digest at room temperature with 35 cc. of dilute nitric acid (specific gravity, 1.20) with frequent stirring. Filter the acid solution through the second filter which was not removed from the funnel, and wash the beakers and filters thoroughly with hot water. Neu- tralize the filtrates with ammonium hydroxid, slightly acidify with nitric acid, add 10 grams of ammonium nitrate, dilute to about 125 cc., and heat on the water bath to 60°C. Add 100 cc. of acid ammonium molybdate and continue the phosphorus determination as usual. MAGNESIA MIXTURE METHOD FOR INORGANIC PHOSPHORUS IN EXTRACTS OF ANIMAL TISSUES. Measure out the number and volumes of extracts specified in the schedule of determinations. Add 10 ec. of magnesia mixture, stirring freely; allow to stand 1916] FORBES AND WUSSOW: INORGANIC PHOSPHORUS ESTIMATION 225 15 minutes, add 25 ce. of ammonium hydroxid (specific gravity, 0.90) ; cover and allow to stand overnight. The next morning filter, and wash the precipitate with 2.5 per cent ammonia water. Dissolve the precipitate on the filter paper with dilute nitric acid into the same beaker in which the first precipitation was made, and wash the papers thoroughly with hot water. Render the resulting solutions nearly neutral; add 5 grams of ammonium nitrate; heat to 65°C.; add 50 cc. of official acid molybdate solution, and keep at 60°C. for 2 hours; continue in the usual way for the gravimetric estimation of phosphorus as the pyrophosphate. NOTE BY H. S. GRINDLEY. The extract of the muscle was prepared by the centrifuge method as used in this laboratory rather than by the method as outlined in the directions sent out for the referee work. This was done merely because it was more convenient at the time. The barium chlorid method was not used on the blood because of the presence of the ammonium sulphate. The results on blood are poor on account of some un- known error in our work, We did not feel justified in making up another solution of blood for the repetition of this one determination. No difficulties other than those usually attending the preparation of solutions of animal tissues were encountered and the details of the procedures connected with the three methods were fairly satisfactory. DETERMINATIONS ON PLANT SUBSTANCES, 1912. Triplicates on _ solutions a-10 gram sample. Gorn sean prepared as directed in b-10 gram sample. Wheat germ the following paragraph. | c-10 gram sample. Rice polish A Wihcatiean Triplicates as above but a-10 gram sample. with 25 ec. of phosphate b-10 gram sample. solution added to each. c-10 gram sample. COLLISON METHOD FOR INORGANIC PHOSPHORUS IN PLANT SUBSTANCES. Weigh out 10-gram portions of the samples in triplicate, and place in 400 cc. Florence flasks, to which add exactly 300 cc. of 94 to 96 per cent phosphorus-free alcohol, containing 0.2 per cent of hydrochloric acid (0.2 per cent actual hydro- chloric acid) and close with rubber stopper; shake the flasks at intervals of 5 minutes for 3 hours, and filter through dry double filters into dry flasks; measure out 250 ce. aliquots of the filtrates into 400 cc. beakers; make just alkaline to litmus with ammo- nium hydroxid and allow to stand for 8 to 12 hours, or overnight. Filter through double filters, and wash with slightly ammoniacal 94 to 96 per cent alcohol. In case a small portion of the precipitate resists transfer from the beaker by the usual means the last traces may be dissolved in 5 drops of hydrochloric acid, with the assistance of a rubber-capped rod. To this acid solution add 10 cc. of alcohol; make slightly alkaline with ammonia, and then transfer to the filter. Wash several times with ammoniacal alcohol; then spread out the inner papers with the precipi- tates and allow to dry completely. Transfer papers and precipitates to Erlenmeyer flasks containing exactly 100 cc. of 0.5 per cent aqueous solution of nitric acid (0.5 per cent of actual nitric acid). Close the flasks with rubber stoppers; shake until the precipitates are thoroughly broken up, and let stand overnight. Filter through dry double filters into dry beakers; pipette out 75 cc. of each filtrate and determine ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 226 t¥0'0 | get0'0 oni a our Sree aguaay 9840°0 Naas 19400 £10 0 ee 8190°0 go 083 7-9 8840°0 cS ee 1940°0 60600 Pe ae #1S0 0 9G 093 g-9 * O110°0 £210°0 mee eas 2°10°0 G1Z0 0 0 00¢ (Gui) o110'0 | ¥eT0'0 | “*" ree 9910'0 | 01Z0'0 0 00g «| 1-9 poyjow oingxrur UIsoUsEyy ase Ren ozto'0 | 2900°0 eneehe Me awh Sees ogmIOAW £1h0'0 6110 0 yee P9100 14 06 ¥-4 Zer0 0 1210 0 9% 0S@ €-4 100 0 Se 0 00¢ 6-4 1700°0 am 0 00g | 1-4 poyjour prroyyo wnrsv¢ Atitbon avo nee 28800 nich shee ggnroay 1¢£0'0 92¢0°0 16£0'0 14 093 ¥-8 CHE ff Sete 0640°0 gL£0°0 Go 06a £-8 ate 0S10°0 68100 ee gone 0 00¢ g-8 Seay oft0'o | ooo | o" pegs 0 00g 1-8 poyjour oyepqAour [v1INAN ¥ poolst eat dedaa0 iP moons Sa000 6°96 90#0°0 fe Sees gmaay 2 eel £280°0 Nie ea O1F0'0 4 096 iat LOPONO is) Meee eee OKO0Ome ter l7On On tn runes 9180°0 Bete: 20400 &G 086 £-9 ee Sh 0 i 6080°0 ae Seok: 0 00g 6-2 ae Geb’ 0 orgs0'0 fo ES 0 00g =| 1-9 (sequtoosse pus S9q10,7) POTJOU oANy xT BIsoUDTYY c FP doconbe lll) Soetma= |b ononne |] osn000 #8F0°0 one oe oa RIOAY Sich. ins £780 '0 £840 0 9% 096 Less mene COSO Os OT PONOK: | cruunen 1€80°0 cero 0 9% 086 £-q OFF O or0g 0 F180 0 pe oh a 0 00g o-4 ObPF 0 OLTS 0 ¥£80°0 é ines 0 008 sl (pyemosurg pay porsy -3019) poyjour pworyo wumntuieg Sd0. 44) socoedoss |) “odddoo =| | ccodooos | lenAacue oa Pelosi's tafepent esvIOAY Sen |S ERONONe | ee SEOi ON |e varius $60°0 14 096 as veo : cee Z180°0 So 096 ie osTs 0 1g80°0 0 009 one es ocrg"0 | 9280°0 0 00g Tt (Ess [seas f3) 191883 JOU) poyjzeur ayepq Ajour [BIINENy sg OPOSN IY unio 1.99 sad uni6 uno qua sad uni quao sad wo16 quao sad UuoLs “99 29 Es 5 FS Es ey 5 iS iy = | $ ig rae tof ny 35 g0 agne 35 SUR as ° ocl = © S09 oun ° a 5 5 sae | s& | Bes | 83g | ow ae Seok ge | bee | be q <6 5 a& 228 < 26 age 5s0 e5 ag f B48 Spe | &2 | SEE | soe | Ge Be Bigiae Be) See eee S ok a BS | £08 9 eg Fees 2 ES aa 3 go go @ "GB he) ° Ego 2° ® 22 ss NOILVNINUTLAG oe. ade ao a. pe oe Pisa ey Be ere 4 ANY dOHLaK ‘GoNvisHos re ea od Hate inl =3 cee Sa oo 36 2 a6 | as r¢ 36 | Bf hs 786 5 i) ag z oe es ssou “1° (LSATVYNV ‘NVOOH ‘“D ‘y¥) jee ig a GNV AWIGNIUD ‘8 'H AD AIMM@MOML “A *d EN Ie SEIN aE 8 ((99 Q00‘¢ 09 dn opvur syoR14xq) ‘(@161) saounjsqns powiun ur snsoydsoyd aunBuour fo uoyounusajap fo spoyjzau fo 4897, T ATaVL 227 INORGANIC PHOSPHORUS ESTIMATION 1915] FORBES AND WUSSOW “‘SUIBIS (OT ‘S8OY put Ao[putixy Aq ‘sures COT ‘JTeo plo sAep ¢ ‘eBpruqamozy, Aq ‘suvId gezo OOT ‘jJ¥O Ysery ‘MossnA, Aq pasn uteI_ ¢ *“poo]q Yselj JO SUITIS G'08Z 0} But -puodsel1109 poolq paze] exo Jo SuIeIZ ONE ‘B80Y] Puv AaTpuTIN Aq ‘suIBIS 7707 GFE ‘JIEO PO SAEP ¢ ‘aBpruqmory, Aq ‘sureIs gF97'9/8 ‘Std yseay ‘MossnAy AQ posn poor” » ‘SUIvID (¢ ‘pelIp WINNowA ‘ss0x pue Ao[puly Aq ‘suelZ cf ‘polmp wNNowA ‘ospriqmory, Aq ‘sweBls 77g/"Cp ‘pelIp WINNOBA SEA MossnA, Aq Pasn aposny ¢ “W813 OTFO'O JNOgGe useq 9ABY P]NoYs ; *WIBIS §[F0'0 JNOGE udeq BAe Pynoyg + Sriferaters eoteieys Taleie: de Ganesc 61F0'0 ac one eiattT arate + -o8pi0ay a : 070 0 14 ee £ ‘ K ie i 81#0°0 14 rea g Amiens pea ae ae Fy 02400 S3 << if PUPS BISeUsBI YIM UOIZezIdweld Yoong ; ae 1z#0°0 On rap teehe sees s QRRIaAy 6160 0 14 be £ ; - Té0 0 14 BES z oo Z2#0 0 of oR I POy}EUL GINgXtI vIsausE ly bes0n0 ez¥0'0 36 vite Dresses ggIOAY ‘ Fe co | eaters | 6240 0 So ake £ opus, ||l\anoote JI ‘goon : #2#0'0 oz vee z ae . tees 0z#0°0 0% sels if poyyeur projyo unwed sdhoap al. Gooude =| \snocan > CZ¥0'0 fo N96 Persse sess ORBIBAW $240 0 93 aa £ ; 9240 0 So oe G me *2F0°0 %% 74 if poyyeur ayepqAyjour ;erynaN :WOTgN[Os eyuydsoy 7 IO0900 96£0'0 walsivieve Te eo. |boaesddcod “oger0Ay 62900 10¥0'0 2060 °0 7, 08% tee 0490°0 26£0°0 8660 0 Ose £0 £080 0 in St 01200 00¢ o-9 0c£0 0 9120-0 00g Ia poyyeur oINjIXIUM BIsoUsE]y Sts eins 9FI0'0 MinretaGats As ress Og BIOAW 9110 0 a) 0L10 0 SS 086 #-q S110 0 0F10 0 8S10 0 SS OSG &-4 S00) 0/5 i) is eerie |e oa 8£00°0 0 00g o-4 29000 eee ¥£00 0 0 00g seal POUPEUK PHO [youre sy, micheal ¢70'0 Aeaeifel= no ASE ese a gRIOAW &L90'0 0820 0 £890 0 &% OSG ia 2990°0 1820°0 0FS0 0 9% 08% £-8 Te£0 0 2360°0 0 00S (rtd 42800 0820 0 0 00g ee poqzeur oepqAour [e1ININ oh UIBIT “ponurydoj—|T ATAVL “WO}BUYMIEZap IB]NFaz Ut v9xH} Jou yonbiys ur 4yo] AyUBND soy seoUBMOTLY FurywUr 19938 ‘paute;qo aq prnoys yorqas AWUUNb Jo ssaoxe ur snr0qdsoyg ¢ “UIB13 YT10'D NOG’ ueeq EBABY P[NOY, ; “snioqdsoqd jo urnsd £000" eAvd osvo YOvE UI YUB[q ey, *UIvIF ggTO'9 OBBIAT INIT Z810'0 ‘g e[dweg :uBI2 gg10'0 ‘7 e[dureg ‘snsordsoyd JO WIBIS 98100 ‘I a[dureg ur OAvS ‘dd Cg UT eIN]XIMA IsaUZBU YIM A]}OoITP pazuztdioard uorqnyjos oyeydsoyg y X th Alias = ic a ; 2£0°0 “1° *! 0810°0 eee rhe | taste ae G98 0010°0 | 1880'0 | “***** |""*|°""** ‘edur8Ay Ss GFIO'O |S | $3 GOTO) Sis | eet) ee Seen | een “| """"* | 0£00°0 | 0900'0 aes 6600°0 | °°’ *" | #810°0 | 9 = OSTO 0 |S@ | 2 00c0'0/9e |} ®@ | | ge | eco ~ | 8800°0 | 69000 | °*° 40100} 68100 | ¢ « 8210°0 |S% | P 0610 0\e% | P aa |p ey ae "1 $£00°0 | $900°0 be Sy 1600" 0} °'°"** | 88100 | i: eS ¥080'0 | 8900°0 }0 | 9 0800°0)0 | 9 | ¢280'0 | €400°0 | 9 | gzto'0 | 8200'0 | Zt00°0 : “| €280°0 | #800°0 | & Ss “""* 1 #0800 | 8900°0 }0 | 4 0800°0/0 | 4 | 8€80'0 | 9200°0 | 4 | £210'0 | 8200°0 | z#00°0 Boe It sae ¢280°0 | $8000 | @ Ss pearance G220'0 1900'0 0 | ® FL00° 08 ® | 2600 | 8200°0 | ® | TTTO'0 | ¢200°0 | OFO0'0 | ~ “"""* | 6680°0 | 88000 | T UBIC BOTA — coud || nee igddve | son|seollGondanedl| cadap: llezpearaio e900} occ + geo-0 =) | g-¢e | goroco | e1z0'0 | + agnraay Z 9110'0 eee 0S10'0 (Td J tee aivleis = [| “a)s.0 siete . oe shee . . ae . . F000 eee . on 9 D 8210'0 | “°°*"* | G910°0 |&% | ® |STT0'0 Vise 09T0'0\¢z | ® is 5 ogee hig Pee g = T210'0 | “°° *** | $8100 |&% | P |2010'0 : 0st 0/¢z | P te eet |e | eee hc LOORO) BBE00KO) ¥ a “=""" 1 FPIO'O | #00" 0 ) 1) CNG), Il) Sees FS10'0)S% | 9 | 0910°0 | 9£00°0 | 9 | 29000 | 100°0 | Zz00°0 £ iS) See nn | ge eater geese | Temes alt nan Le tt FO 1 020-0 { 8F00°olo q | S910°0 | 4800°0 | q | $#00°0 | 0100°0 | 8100°0 z S a 0920" Op AAC LORO 8 (Ole ee If \ cF00'0)0 | B | S910°0 | L800°0 | & | €200°0 | 21000 | 02000 I ysrjod eany Sferetetels Snege |! manors 50 ol onpdioe Os 50% reesesdeesleoel ppegeg | ceeees [eeel grrgcg | coeces | cere ee sede 9FBIIAY z See lee eal ao lee reer tal aces terre ol oy | asccecel enecs | Mall esate Necones lle eoell erate - a eae ca Wk Se cote | | eh "12010 °0 : 8810'0/S | 3 et || een | cee | mete WEEOORONEB900%0) PO TARO) Wp ee #020'0 | 9 ES 9400°0 “* | 9810°O |S@ | 9 |2z10°0 4 £020'0/9 |} 9 | ~"""* Seen | ana peamne ati OS00ROMINOLO0K0 : 9110'0 | ~* 9020°0 | ¢ tS) S110" OU ee £610°0 |S@ | P |STt0 0 aos 9610°0|S2 | Pp eae || penned | ucts | eee LE OORON (W201 0, vi FIIO'O | °°" | #020°0 | & f “* | G8€0°0 | 8200°0 |0 | 9 |SeT0°0 me 9120 '0/&% | 2 | 88£0'0 | 9200'0 | 2 | ZF10°0 | €£00°0 | 8F00'0 oye “"""* 1 10F0'0 | 0600°0 | & o “es"* 1 9€80°0 | #40070 10 | 4 eaesll 920°0 { 0800°o\o q | 8880°0 | 9200°0 | q | 8600°0 | 2200°0 | 2800°0 g “""" 1 9040'0 | 1600°0 | < “"*="* | 00%0"0 | 8800" 0 Ore 35 cee J | 2800°0}0 | 99800 | 0800'0 | @ |} ITI0'0 | $200'0 | 0F00'0 fs gs |wle| wy SB |e] 9 5 = >| WF ale < Ge les | ee [S|6| Se | ee | ee ele) & | ee 1b) 23 |eee) gee ieegel Se | eee (6 =} p03 ae) 6 \o oo S09 i=} DB hig ue Bolg no boop BSE B260| Of 5905 |0 3) eles ee eo |SI/=| Ba a8 Ded cove | PON | hoa} | Bo Ve | fs O65 5 @ eeogal as o2o |= 5 3 25 | o8 jelo| oo | 85 | SB jejo| & of |o| gs | oo 8 29 &. Sesc0l £8 | ac2/° Pas =p oe |2 a8 = oc |& i} OF oR Pe Pe Bee 2s BE z ER 3, | SH |° 8 3 | Se | 8 Eas a a a5 yas! So Be < Os = Bs 2 o5 > Big Z zs Big 1 or Spo on IS rae i Be | 22 9 Be o/c > 52 Se Sir) 2.2 $3 ge s) Bo |¢. ge 3 ca ee Bi gid 2pezoel} "3 SBS! 6! acs SON YLsans <5 Be ty ey 25 a eS |S E. *S -xo snioyd g Be c| a u ; £5 3 Bee (pes 5 Be 5 Ea -soyd Jo ssoaxny neh x ag & a WE ae a HWE a 4 nN ne = = ar) 2. = : ca ISU 7 Ey u T g NOIMWOVULXG GNOOTS NOILOVULX (ISA TVNV ‘NVYDOH ‘D ‘y) GIHA *4 ‘H ssou ‘I ‘ad NV TMOSSOM ‘a ‘a‘y GOCINAMOUL “A ‘ad AGIANIUS 'S‘H “(GI6]) saounjsqns junjd wr snsoydsoyd avunbsowr fo uoyounusajap fo spoyzau fo 48a 7, % ATAVL 1915] FORBES AND WUSSOW: INORGANIC PHOSPHORUS ESTIMATION 229 phosphorus in the usual way, precipitating first with acid molybdate solution, then with magnesia mixture, and weigh as the pyrophosphate. If the final solutions are highly colored, dissolve the pyrophosphates and reprecipitate. NOTE BY H. S. GRINDLEY AND E. L. ROSS. The details of procedure worked satisfactorily. It seems to us that if possible the weights of magnesium pyrophosphate should be increased in order to insure more reliable final results. We regret to say that as a result of a number of experiments we are convinced that the acid alcohol method as outlined does not insure complete extraction of all the inorganic phosphorus from meals, or the solvent gradually converts the organic forms of phosphorus into the inorganic form. A second and a third extraction of the samples with the 0.2 per cent acid-alcohol, after making the proper corrections in each case for the aliquot portion from the former extracts, gave additional in- organic phosphorus. The second extraction gave the following percentages of phosphorus: Corn germ, 0.0052; wheat germ, 0.0012; rice polish, 0.0018; and wheat bran, 0.0011. Proceeding in a similar manner the third extract gave the following percentages of phosphorus: Corn germ, 0.0004; wheat germ 0.0032; rice polish 0.0035; and wheat bran 0.0029. It thus seems that the third extract apparently contained more organic phosphorus expressed as per cent than did the second extract. This is probably accounted for by the fact that the filtration of the acid alcohol for the third determination was made during the middle of a very hot day and the filtration proceeded slowly so that a considerable loss of alcohol by volatilization resulted. In fact we were unable to get the usual 250 ce. portions from the filtrates and had to take only 200 ce. for each determination. This loss of alcohol by evaporation would result in giving higher results than should be obtained. Other tests of a similar nature gave similar results. Further, dipotassium hydrogen phosphate deposited as a solid in the tissue of filter paper apparently cannot be completely extracted by 300 cc. of 0.2 per cent acid alcohol under conditions such as those now used in this method for the solution, separation, and estimation of phosphorus in vegetable substances. DETERMINATIONS ON ANIMAL SUBSTANCES, 1913. f a-1 Neutral molybdate precipitation. a-2 Neutral molybdate precipitation. A. Extract ofsamples} a-3 Neutral molybdate precipitation. as weighed. a-4 Magnesia mixture precipitation. a-5 Magnesia mixture precipitation. Muscle a-6 Magnesia mixture precipitation. Blood Brain b-1 Neutral molybdate precipitation. B. Extract of samples | b-2 Neutral molybdate precipitation. as weighed plus 25 | b-3 Neutral molybdate precipitation. ec. of phosphate | b-4 Magnesia mixture precipitation. solution. b-5 Magnesia mixture precipitation. | b-6 Magnesia mixture precipitation. Determine by both methods of precipitation the phosphorus in the phosphate solution used; also make blank determinations in triplicate on reagents. 230 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 PREPARATION OF COLD WATER EXTRACT OF MUSCLE. A. Weigh out 10 to 12 grams of fresh muscle and divide as nearly equally as possible between two small beakers; moisten the samples with a few cubic centi- meters of distilled water, and break up lumps with a glass rod; add 50 ec. of water to each beaker and stir contents for 15 minutes. Allow insoluble residue to settle for 3 to 5 minutes, then decant the liquid from each beaker through filters into beakers; allow to drain and add 25cc. of water; stir for 7 to 8 minutes, and after allow- ing to settle, decant onto the same filter. Continue this treatment, using each time 25 ee. of water, until the filtrates measure about 230 ce. each. Allow the filters to drain completely between extractions. Whenever the major portion of the residue has become mechanically transferred to the filter return it to the beaker, using great care not to break the filter paper. After the last extraction throw the entire con- tents of each beaker onto the filter, and after draining, wash twice with small quanti- ties of distilled water. Combine the two extracts, and use for the precipitation of the phosphates under Section A. B. Weigh out same quantity of flesh as specified for A, and divide as nearly equally as possible between two small beakers: work up with a few cubic centimeters of distilled water; add 25 cc. of aqueous solution of disodium phosphate equivalent to about 40 mg. of magnesium pyrophosphate, dividing as nearly equally as possible between the two beakers and proceed as directed under A. The extract thus ob- tained is ready for precipitation under B. PREPARATION OF HOT WATER AMMONIUM SULPHATE EXTRACT OF BLOOD. A. Weigh out 30 to 35 grams of fresh blood, or the equivalent of oxalated blood, into a 400 cc. beaker; add a few cubic centimeters of distilled water, and work up the blood and water with a glass rod; make up to about 150 cc. with boiling distilled water; place over a flame, and gradually bring to boiling, with constant stirring. When boiling begins add 20 ce. of 20 per cent ammonium sulphate solution. Boil with constant stirring, for about 10 minutes; decant onto a filter of sand on linen, receiving the filtrate in an 800 ec. beaker. When the liquid is through, lift the coagu- lum from the sand, and transfer it to a mortar; grind to a smooth paste and transfer from mortar to beaker with boiling distilled water; make up to about 50 ce. with the same; stir for 8 minutes, and pour contents again onto the sand filter. After the extract is through, return the coagulum to the mortar, and grind a second time, transferring to the beaker as before with boiling distilled water. Repeat this process of 8-minute extractions of the coagulum in hot water, and filtration as just directed, without further grinding, until the filtrate measures about 450 cc. Wash out each beaker twice with 8 to 10 cc. of hot water, completing the transfer of the coagulum and extract to the sand. Wash the coagulum on the sand twice with boiling water from a wash bottle. At all times allow the filter to drain well between additions of extract or wash water. This extract of about 500 cc. is ready for pre- cipitation under Section A. B. Weigh out same quantity of blood as specified for A. Work up with a few cubic centimeters of distilled water; add 25 cc. of an aqueous solution of disodium phosphate equivalent to about 40 mg. of magnesium pyrophosphate and proceed as directed under A. The extract thus obtained is ready for precipitation under B. PREPARATION OF HOT WATER AMMONIUM SULPHATE EXTRACT OF BRAIN. A. Weigh out about 10 grams of brain into a 250 cc. beaker; add a few cubic centi- meters of distilled water and work up the brain and water with a glass rod; make up ————— 1915] FORBES AND WUSSOW: INORGANIC PHOSPHORUS ESTIMATION 231 to about 100 cc. with boiling water, place over a flame, and gradually bring to boil- ing with constant stirring. After boiling has begun add 20 ce. of 20 per cent ammon- ium sulphate solution; boil gently for about 10 minutes; allow to settle for a moment, and decant liquid slowly onto a filter of sand on linen,! receiving the extract in an 800 cc. beaker. Add to the beaker containing the coagulum 50 ce. of a 0.1 per cent ammonium sulphate solution; stir for 1 minute, keeping over flame and at the boil- ing point; decant the liquid onto the filter. Repeat this process of 1 minute ex- tractions of the coagulum in 0.1 per cent ammonium sulphate solution, and filtration as directed, until the filtrate measures about 450 cc. Wash out the beaker twice with 8 to 10 ce. of hot 0.1 per cent ammonium sulphate solution, completing the transfer of the coagulum and extract to the sand. Wash the coagulum twice with the above wash solution from a wash bottle. At all times allow the filter to drain well between additions of extract or wash solution. This extract of about 500 cc. is ready for precipitation under A. B. Weigh out same quantity of brain as specified for A; work up with a few cubic centimeters of distilled water; add 25 cc. of an aqueous solution of disodium phos- phate equivalent to about 40 mg. of magnesium pyrophosphate and proceed as di- rected under A. The extract thus obtained is ready for precipitation under B. NEUTRAL AMMONIUM MOLYBDATE METHOD FOR INORGANIC PHOSPHORUS IN WATER EXTRACTS OF FLESH. Treat 3 of the extracts prepared according to the directions for A, and 3 of those prepared as for B as follows: Evaporate, with frequent stirring on the water or steam bath, to approximately 20 to 25 ec.; while hot, filter into 300 ec. beakers, using doubled 11 em. 8. and S. No. 589 ‘Blue Ribbon’”’ papers. Wash beakers, precipitates and filters thoroughly with hot water. The volume of the resulting filtrate and wash- ings should be about, 125 ec. Add 10 grams of ammonium nitrate and heat upon the water bath to 60°C. Then add 10 ce. of nitric acid (specific gravity, 1.20), stir, and add 125 ce. of clear neutral molybdic solution. (Neutral ammonium molybdate is prepared by adding ammonia to the ordinary molybdic solution, using litmus paper as an indicator. This work should be done very carefully and both red and blue litmus paper used.) Reheat, bringing temperature to 60°C.; keep at this tem- perature for 15 minutes; stir vigorously every few minutes during this time. Re- move from the bath and allow the solutions to stand 2 hours in a warm place. De- cant the clear supernatant liquid through doubled 11 em. No. 589 (Blue Ribbon brand) S. and S. filters. Transfer the remaining liquid and precipitate to the fil- ters, using a 10 per cent ammonium nitrate solution. Wash precipitates and beakers four or five times with small volumes of the ammonium nitrate solution. Dissolve the yellow precipitate upon the filter and that in the precipitating beaker with dilute ammonium hydroxid (2.5 per cent) and hot water, collecting the filtrate in a 250 ce. beaker. Wash thoroughly; neutralize the solution with nitric acid (specific gravity, 1.20) and make up to approximately 150 cc.; add 5 grams of ammonium nitrate; heat upon the water bath to 60°C. and then carefully add, while stirring 5 cc. of concentrated nitric acid and 50 cc. of clear acid molybdic solution. Digest at 60°C. for 15 minutes, stirring occasionally. Continue the determination of phosphorus as usual weighing the phosphorus as magnesium pyrophosphate. * Tt is desirable to prevent the extract or coagulum from coming in contact with the linen before passing through the sand. To this end pour extract slowly onto center of sand or into a cup-shaped depression. 232 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 MAGNESIA MIXTURE METHOD FOR INORGANIC PHOSPHORUS IN EXTRACTS OF ANIMAL TISSUES. Treat 3 of the extracts prepared according to the directions for A, and 3 of those prepared as for B as follows: Add 10 cc. magnesia mixture, stirring freely; allow to stand 15 minutes; add 25 ec. of ammonium hydroxid (specific gravity 0.90); cover, and allow to stand overnight. The next morning filter and wash the precipitate with 2.5 per cent ammonia water. Dissolve the precipitate on the filter paper and that remaining in the beaker in which the precipitation was made with dilute nitric acid (1: 1) and hot water, receiving the solution in 400 ce. beakers. Neutralize the nitric acid with ammonium hydroxid; make slightly acid with nitrie acid. Add 5 grams of ammonium nitrate, and precipitate in the usual way with molybdate solu- tion. Continue in the usual way for the gravimetric estimation of phosphorus as the pyrophosphate. Blanks and phosphorus estimations on phosphate solutions used in work reportedin the preceding table. SAMPLE BOCUDIONE, aeaeeeaited | gram Blood: Blank 1: Neutral molybdate precipitation............. 0.0020 Blank 2: COKE A 04 Oud me ene aatarteeeteC rien 0.0016 Blank 3: Ch eM aoe hacde taritncs 0.0018 Average 0.0018 Blank 4: Magnesia mixture precipitation................ 0.0030 Blank 5: COP Se ee os Stee eerie 0.0037 Blank 6: GOP a" Pare Wee encan ee eae 0.0038 Average 0.0085 Brain: Blank 7: Magnesia mixture precipitation............... 0.0010 Blank 8: Goi ie eel eae evecnnscens 0.0010 Blank 9: Clem Ne Piles ¢ Anaernesurn cons 0.0006 Average 0.0009 Blood and muscle: Phosphate solution 1: Neutral molybdate pre- cipitation 25)Gce 0.0419 Phosphate solution 2: do 25 cc: 0.0416 Phosphate solution 3: do 25 cc. 0.0418 Average 0.0418 Phosphate solution 4: Magnesia mixture precipi- tation 25. CC. 0.0417 Phosphate solution 5: do 25 cc. 0.0415 Phosphate solution 6: do 25 cc. 0.0411 Average 0.0414 Phosphate solution 7: Direct precipitation with magnesia mixture 25 cc. 0.0413 Phosphate solution 8: do 25 CC: 0.0419 Phosphate solution 9: do 25 ce. 0.0419 Average 0.0417 Brain :! Phosphate solution 10: Direct precipitation with magnesia mixture 50 ce. 0.0266 Phosphate solution 11: do 50 ce. 0.0269 Phosphate solution 12: do 50 ce. 0.0264 Average 0.0266 Phosphate solution 13: 50 cc. diluted to 500 ec.: 163 grams ammonium sulphate added 0.0264 Phosphate solution 14: do 0.0262 Phosphate solution 15: do 0.0267 Average 0.0264 1 For modification of method used with brain see p. 236. 1915] FORBES AND WUSSOW: INORGANIC PHOSPHORUS ESTIMATION 233 RESULTS ON ANIMAL SUBSTANCES, 1913. TABLE 3. Test of methods of determination of inorganic phosphorus in animal substances, 1918. (Analyses by A. F. D. Wussow.) o8- rronus |aacyesrow| worcantc | ,$PBEP PHOS SUBSTANCE, METHOD AND WEIGHT OF ADDED AS PYRO- PHOS- ERED AS DETERMINATION SAMPLE MAGNESIUM| PHOSPHATE] PHORUS MONON ETT PYROPHOS-| OBTAINED OBTAINED PYROPHOSPHATE PHATE grams gram gram per cent gram | per cent Muscle: Neutral molybdate method. IGE OM Soscoc 0.0276 | 0.0568 |...... MN RN os oui 0.0240 | 0.0578 |...... 13% 3025|ieeeeee 0.0259 | 0.0543 |...... AC ELA POUR ee etter call © o ieischeaurelll aeretecill | mremiente OF0563)| fencer 1189305 ieee er 0306481) |e eee 0.0407 I AGBYO) “oesues OF0636))|eecnee 0.0403 1226880] eee 006547) erence 0.0898] .... AWOL ADO Meet aceretisesiet|). acu ates OL OAL Til) Soyo meee 0.0403) 96.6 Magnesia mixture method.| a4 14.5075] ...... 0.0303 | 0.0582 |...... Bats 1OFO850| eee 020219), 020605) ||. .=--. LATS OT eae eee 0.0229 O205595|Feeere PA ORAL OP te eiayi ccs atey All ca. eles Siebel eo) mei kastaten| ueetee es OFO5825 Renner DUE ZOGS|) eeecscuorene OROG1 Zi eee. 0.0373 1 3530| eee MCG | sosdse 0.0369 2045S reer OROG2IG IT Raeeae 0.0368) .... PAVCLOR OM ncctectiasrndsaest ). sastarete CUO SEAN is toate o goal li ne ae 0.0370] 88.7 Blood: Neutral molybdate method. 291950 reese 10.0324 | 0.0309 |...... 3520955| secre 100359) || 0.0285 |...... 23) 4170 |e LF O25 7s |MMORO25 2) |erreree PAV CTAR CI ices sisieieisie cs e-akll © aa PRO RII pie cee enellln Ney ee O202825i tear 36.5820) ...... AD SOAES: | gece. 0.0348 PUREE Gaoocs DOG || onosos 0.0385 32.8275). 2... AD AUG) || pecoen OOBGYA) Cone INV ALG oases cle ce bie,ccial |e glare OOS Te eee, ll tances 0.0367) 88.0 Magnesia mixture method. 3059601 eee LOROUS (a eOLOIAIG i eeeer acs 32.4669)... 5... 10.0146 OFOI25) Reeere aton'd DOP2820 FT ores. LORONSI | OR OV4Ay eee secre AVIELAL Cecio Sebeas diac |e ys onene Neri eetaey: al dom een OFOUS 7a eee Ae BOR OST ein erie LOR OD45 eee 0.0370) .... SePGMI|| Saceor LORO5G5)||eenae 0.0392] .... 3O ele) Vee. 10% 05327 | erect 0.0354] .... AVC AS Crap rteourscinase ac] i Seeeeene OR OAM Rieter ceasl oe arena) 0.0372] 89.2 Brain :2 Magnesia mixture method. LOS1650|; wee. 10.0246 | 0.0675 |...... 1023040) .y34: 2002510) OS06790heacnc LOES550|\" Seen 10.0267 | 0.0679 |...... JANSEN OS aero eraterr ere el || ete toes hcl Nie eee Ie Minced OROG7S5 | Rees 10.5095) 0.0266 | 10.0512 | ...... 0.0257) . 9.2880} 0.0266 | 10.0483 | ...... 0.0257) . 10.5465} 0.0266 | 10.0514 O20258i) ee. 0.0257] 96.6 Jal ER ead ec 1 Blanks deducted. 0.0266 * For modification of method used with brain see p. 236. 234 TABLE 4. ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. J, No. 2 Determination of inorganic phosphorus in pig’s blood extracted with 33 per cent (Analyses by A. F. D. Wussow.) ammonium sulphate solution (each extract about 500 cc.) MAGNESIUM | pemmmursa- | weiomr OF | enearawr or exrnact cee eee OBTAINED grams gram F per cent Lee oe rare 33.0990 | Evaporated, boiled, and filtered| 0.0217 0.0183 Deed on orate 34.4230 do 0.0222 0.0180 ‘Average. .<|l voce nose oe Oe ee | Re deecycescane 0.0180 Spears sion 33.3960 | Evaporated and filtered 0.0216 0.0180 yn aatnae 31.3340 do 0.0218 0.0194 Averare::.|| (Sits 720 1 |e Prey pean meer BOO Fill al ee 0.0187 Peete saben 33.7026 | Precipitated directly 0.0108 0.0089 Gi. Siete tases 33.2960 do 0.0108 0.0090 Averape:s({|-ee See le ie tl eee ee Bei Bro) xyay ey 0.0089 TABLE 5. Inorganic phosphorus in brain determined by modified method. (Analyses by F. M. Beegle.) PHOS- > 7. = 4 R Sonepat, tenon) wed | Pe@enc [ACER [aroncamnc |, FORE | Aon eae N > ae y NE! - SEE cance | | ORES) wzomve PACU | Ao PHATE grams gram per cent gram gram per cent Brain :1 Magnesia mixture method ee rari |) Oe | OUST Seaceo |lboooe< a-5| 9.2368 | 20.0209 | 0.06518 | ...... |...... b-4 | 10.7215 | 70.0507 | ...... 0.0267 |0.0263 b-5| 9.2277 | 20.0474] ...... 0.0267 |0.0264 b-6 | 10.5182 | 70.0502 | ...... 0.0267 0.0263) .... AV ETE BO arc saele soluvenis ellis Sireyelonl weer cbt Sees 0.0267 |0.0263} 98.9 Phosphate solution (50 cc.): . | ee cree (OTT seo. ell Loaoadallscdcos | DNR cates O02) Moe ae Iie geen goo sc | B})| omeene (OPH SGaaade vl) eaosdeo |basoos AVEFA Ge sack tenh «as ees unos (OOP AM I Stace iyesoaeagn|sonocc Blank: ThE || aac OSUOU | seacos |) bocctd |[sonose Dl). cpeeerege OXOO0G HI! Aes mil ee Seescal| eee Sidecars OF0008 | Wesees Mil) Aes eee AV ETAT Cosa ties ceenee (OAC eecoee | Ile aoeeorn ietncac 1 For changed details of method followed see p. 236. ? Blanks deducted. 1915] FORBES AND WUSSOW: INORGANIC PHOSPHORUS ESTIMATION 235 DETERMINATIONS ON PLANT SUBSTANCES, 1913. f a-1 10 gram samples. a-2 10 gram samples. a-3 10 gram samples. Triplicates on solutions b-1 10 gram samples—50 cc. of prepared as directed in phosphate solution B. following section b-2 10 gram samples—50 cc. of phosphate solution B. b-3 10 gram samples—50 cc. of Bluegrass phosphate solution B. Alfalfa Rice polish a -1 3 eee eee ataink c-1 10 gram samples c-2 10 gram samples. Triplicates on solutions c-3 10 gram samples. prepared as directed on d-1 10 gram samples—50 cc. of page 232 (aqueous hy- phosphate solution D. drochlorie acid-phenol d-2 10 gram samples—50 ce. of extract) phosphate solution D. d-3 10 gram samples—50 ce. of phosphate solution D. Determine strength of phosphate solutions B and D; also make blank determi- nations in triplicate on reagents. A. AQUEOUS HYDROCHLORIC ACID EXTRACTION. Pour exactly 300 cc. of 0.2 per cent hydrochloric acid (4.6 cc. of concentrated hydrochloric acid, specific gravity 1.18 to 1.19, per liter) onto 10 grams of sample in a dry 400 ec. Florence flask. Close with rubber stopper, and shake at intervals of 5 minutes for 3 hours. Filter the extract by suction into dry flasks through S. and S. No. 589 ‘‘Blue Ribbon”’ papers, in a Witt filtering apparatus, or a Biichner funnel. Measure out a 250 cc. portion of this filtered extract, and precipitate in a 400 cc. beaker with 10 cc. of magnesia mixture and 20 cc. of ammonium hydroxid (spe- cifie gravity 0.9). Allow to stand overnight, and filter through double S. and S. No. 589 ‘“‘White Ribbon” papers, taking care to decant as long as possible without pouring out the precipitate. Then complete the transfer of the precipitate to the paper. Wash three times with 2.5 per cent ammonium hydroxid, and then three times with 95 per cent alcohol. Allow the precipitate to drain, spread out the inner paper on the top of the funnel, and allow the alcohol to evaporate. When practically dry, place this inner paper with the precipitate into an Erlenmeyer flask; add 100 ec. of 95 per cent alcohol containing 0.2 per cent of nitric acid; close the flask with a rub- ber stopper and shake vigorously until the paper is thoroughly broken up. If the precipitate is flaky, and refuses to break up on shaking, allow to stand in the acid- alcohol overnight. Filter through a dry filter into a dry flask; pipette out 75 ce. of the filtrate into a small beaker, and evaporate almost but not quite to dryness. Dissolve in dilute nitric acid, and filter if necessary; then determine phosphorus in the usual gravimetric way, by precipitation first with acid molybdate solution, later with magnesia mixture, and then by burning to the pyrophosphate. The result obtained represents 6.25 grams out of the original 10 grams of material and so to reduce to a 1-gram basis multiply by 0.16. 236 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 B. AQUEOUS HYDROCHLORIC ACID EXTRACTION PLUS PHOSPHATE. Proceed as under A except that in place of 300 ce. of 0.2 per cent hydrochloric acid add 250 ee. of the same and 50 ec. of phosphate solution containing disodium phos- phate equivalent to approximately 25 mg. of magnesium pyrophosphate per 50 ce. Make up this phosphate solution with 0.2 per cent hydrochloric acid. C. AQUEOUS HYDROCHLORIC ACID-PHENOL EXTRACTION. Proceed as under A except that in place of 300 cc. of 0.2 per cent hydrochloric acid add 300 ce. of 0.2 per cent hydrochloric acid solution containing 50 grams of phenol per liter. D. AQUEOUS HYDROCHLORIC ACID-PHENOL EXTRACTION PLUS PHOSPHATL. Proceed as under C except that in place of 300 cc. add 250 ce. of 0.2 per cent hy- drochloric acid containing 50 grams of phenol per liter and 50 ce. of phosphate so- lution containing disodium phosphate equivalent to approximately 25 mg. of mag- nesium pyrophosphate per 50 ce. Make up this phosphate solution with 0.2 per cent hydrochloric acid containing 50 grams of phenol per liter. Nore: Make phosphate solutions used in B and D of same strength by weighing out equal quantities of the phosphate, and determine their exact strength by pre- cipitating in triplicate 50 cc. with magnesia mixture, filtering, igniting, and weighing direct. Make blank determinations in triplicate on reagents. RESULTS ON PLANT SUBSTANCES, 1913. Blanks on solutions used in work reported in Table 6. GRAMS OF MAGNESIUM PYROPHOSPHATE BOR UTLON Grindley Trowbridge Wussow! and White and Newlin Smith Blank 1: Aqueous hydrochloric acid solu- tons! a. Sod at eee eee eee 0.0002 | 0.0012} 0.0000 | 0.0011 2: do 0.0002 | 0.0012} 0.0000 | 0.0011 3: do 0.0002 | 0.0006 | 0.0000 | 0.0007 IAVETR EC tach mnccinaeene 0.0002 | 0.0010 | 0.0000 | 0.0009 Blank 1: Aqueous hydrochlorie acid phenol SOMLONS! ene sates eccioe soe 020002) SOf01GI9|| Seeee 0.0000 2° do 0200025 OROLS9R eee 0.0002 3: do O20002)| SO201635)) 22 aee 0.0004 INE padoneodeaccd | OSU OrA) OOM oS occ 0.0002 Phosphate solution 1: (Aqueous hydro- chlorice acid) 50 ce.| 0.0250 | 0.0005 | 0.0426} 0.0139 2 do 0.0248 | 0.0012] 0.0456 | 0.0135 3: do 0.0248 | 0.0012} 0.0432 } 0.0130 IAVETE PONS caeeise ia: 0.0249 | 0.0010} 0.0438 | 0.0134 Phosphate solution 1: (Aqueous hydro- chlorie acid phe- 1410) ))) G40) COscccnaunee 0.0249 OFR0152 | eee 0.01386 2: do 0.0249 OROUS 7S eee 0.0135 we do 0.0249 IGostiy ||\eeeeeeee 0.0125 Average..............| 0.0249 AOE |) Sonne 0.0182 1 In the second set of determinations with alfalfa, magnesium precipitates were allowed to stand an ex- tra day before filtering, and, after filtering, an extra day in acid alcohol; with Samples a-1 and a-3 only 200 ec. of aqueous hydrochloric acid extract was used, but figures given represent 250 cc. as usual. With blue grass in Samples a-l, a-2, and a-3, only 200 cc. of the aqueous hydrochloric acid extract were used, but weights given for magnesium pyrophosphate represent 250 cc. With rice polish, magnesium precipitate was broken up in acid alcohol with stirring rod before filtering off 75 cc. of aliquot. i i i I 1 Should have been about 0.0153 gram. ? Not included in the average. 3 In the second extraction 200 cc. instead of 250 cc. of the aliquot were used. TABLE 6. JouRNAL OF THE ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS . Test of methods of determination of inorganic phosphorus in plant substances (1913). Avouer 15, Vou. I, No. 2. E. B. Forbes and A. F. D. Wussow. _— A. F. D. WUSSOW H. 5. GRINDLEY AND C. 1. NEWLIN H. L. WHITE P. F. TROWBRIDGE AND 0. c. sMiTHt Magnesium pyrophosphate ; Added phosphorus recovered as AUBBTANCR, METHOD AND DETEMMINATION obtain te oP magnesium pyrophosphste* Magnesium : Auded honoris Magnesium soe Magnesium oan py Rune: ayer se yecoversd) maiz agaesium pyrophos- Tnorganic recovered as | pyrophos- Inorganic beae al Firat set | Secondset | Firstset | Second set First net of Reseed anne e aiedr |e pyrophosphate! neeoh phosphorus | magnesium | phate ob- | phosphorus | magnesium of deter- of deter- of deter- of deter- atacinatiqna daterminstlodas - ByTUp noe: tained pyrophos- minations | minations | minations | minations phatet phate — gram gram per cent per cent gram per cent gram per cent gram per cent gram per cent gram per cent gram gram per cant gram Alfalfe Aqueous hydrochloric acid extraction a-l 0.0189 * 0.0219 a-2 0.0170 0.0184 a-3 a te 0.0160 : Avarksel.ciavieveccacde testy meee. : 0.0180 0.0188 0.0803 0.0838 b-1 0.0188 0.0237 pen b-2 0.0222 b-3 0.0225 0.0235 Avorngo..... - ? vapaee 0.0212 0.0236 0.0032 20.5 0.0048 Aqueous hydrochloric acid-phenol extraction | 0.0182 0.0140 sees seeee . o-2 0.0168 0.0189 wabeos 2 o-3 0.0175 0.0149 Avorngo,........ aUaeL Sete rnaree ae 0.0175 0.0159 0.0780 0.0709 d-1 0.0210 0.0282 z d-2 0.0203 0.0241 d-3 0.0229 ‘ AVOTIEO. 6.6 se ce ene pecpusseuteeesiese 0.0213 0.0236 as. 0.0038 24.3 0.0077 Bluo grass: Aqueous hydrochloric acid extraction a-1 0.0228 0.0380 . trees ee (ee a-2 0.0225 0.0379 Heer R aT || lime a-3 0.0224 0.0335 sae ee a ot Avernge. . 0.0226 0.0365 0.1008 0.1628 bel 0.0283 0.0321 b-2 0.0280 0.0208 b-3 0.0255 0.0320 . A . Feet aes! AVOTBGO...ccvcvsccrrcercenscsonsnseseens 0.0250 0.0313 owrvens 0.0030 19.2 —0.0052 Aqueous hydrochloric neld-phenol extraction ol 0.03865 0.0307 soseee By. aa Ra o-2 0.0355 0.0409 ¢cceae ; was o8 0.0847 0.0408 eens PANT| -cceaa | Al bones AVOTOZO, os ccecncncccsecscsvecccccecs 5 0.0356 0.0404 0.1802 d-l 0.0403 0.0554 d-2 0, 0458 0.0548 d-3 0.0498 0.0545 ad dedeee aeee vere AVOIOQOs ciscaseencddnousvcasceberccters 0.0483 0.0549 Rt: 0.0127 814 0.0145 Browor's grains: Ag\oous hydrochloric acid extraction al 0.0025 a2 0.0021 a3 0.0023 AVOTNBO. 6 cece eeb cobs cneegneens 0.0023 0.0103 b-1 0.0160 : F b-2 0.0148 Sopa | | emia b-3 0.0161 $4 VESREM cpadekaandebede\taeceseues Tences 0.0156 0.0.33 85.2 Aqueous hydrochloric neid-phenol extraction cl 0.0012 o-2 0.0010 o-38 0.0013 avaye on Avorago......... ‘ sane ; 0.0012 o¥e DOSE IE coninn 8 1)\ Pimes d-1 0.0162 | BR en eae ee eee BA! iO LOURB:. e[ PY 4c A cks ill OR .e<- dori | ieee et el Oe cerme Le a-3 15) | a lee et lean oT bees 1 lee fee 1 ANOUEO win wes sessvcdoussapssacs on 0.0160 0.0148 94.0 Rice polish: Aqueous hydrochloric acid extraction 0.0038 DORKS Te. er ae) Resse MiP waters 0.0038 0.0026 0.0035 0.0027 Average. 0.0037 0.0030 0.0165 0.0119 ‘OLO078' | ae... 0,0110 0.0075 0.011 OTe PS Seer Pe || eRe rer p* Average 0.0113 0.0077 0.0076 jueous hydrochloric acid-phenol extraction 0.0027 Ae ee GR eee 0.0017 0.0020 0.0024 DA oe ee en Se eee Average... 0.0023 0.0020 0.0102 | 00089 | ...... 0.0052 0.0064 0.0093 DDR liccxaeee || Meee, |. saccs. I DiGene cee DIOLS es] * secseg Bh )) EXTRACT A EXTRACT B 2B < ANALYST Ag eg Oy Za Zig Eales Be. aa | dan | aaitcas |cemee aS * as | oe ° 12) a per cent) per cent| per cent| per cent| per cent} mg. mm. NWO TELEOTIM MERE er miei sinicctan ficinsns Sate 1.51 | 1.85 | 5.95 | 1.48 | 4.93 | 21.0 (?) 1.54 |12.31 | 5.87 | 1.79 | 5.20 | 23.4 oie 1.51 | 1.96 | 5.91 | 1.10 | 4.99 | 22.6 sie (COG) Ss sup eo ReOW COO S E eee eee Mee ERT uaa ay oto Ilietes | aah Zk 1.40 | 1.33 | 4.93 Syed Berets LOR Sales 1.55 | 1.40 | 4.94 Bor ae 20.2 | 8.0 Rudnick and Irainor:.............. 1.66 | 1.80 ; 6.63 | 1.90 | 6.64 | 21.3] 7.8 1.80 | 1.76 | 6.65 | 1.93 | 6.57 | 19.5 | 8.0 1-87 || 1.88) | 6:62 | 1-91 | 6-60) || 19:4 |) 729 RSLETLU Pe Vases yeast. ais ciclcFevartioveusys le ite 1953) 1522)) (15256: se erate alle OW Se 0) 1.59) |) 1.27 | 5250 Bevo] | ese | ok) abe 1.55 | 1.19 | 5.65 eT ginda | eae ae Emmett and Davisson..............} 1.72 | 0.64 | 6.58 | 1.18 | 5.66 | 21.7} 8.0 1.66 | 0.70 | 6.64 | 1.20 | 5.69 | 21.4] 8.0 1.59 | Lost | 16.90 Se Oe AOn ez Lele || Ze 9 V. C. Myers and M.S. Fine, Post- | 1.68 | 0.69 | 6.35 | 1.40 | 5.97 | 22.4] 8.0 Grad. Med. School, New York, | 1.68 | 0.70 | 6.36 | 1.23 | 5.95 | 22.4] ... Nee 1.68 | 0.71 | 6.380 | 1.83 | 6.01 | 20.9 Summary—Average of triplicate determinations. MUO UIT OTE are yest eit shcs cane oiehs 1.51 | 1.91 | 5.92 | 1.46) 5.04 || 22.3 (7) (Cale. Hae p PUL ete AOS Oe Sen eee 1.46 | 1.33 | 4.88 | 1.23 | 5.44) 19.9 | 7.9 Rudnick/and Drainor.+............ 1.78 | 1.81 || 6.63) |) 1291 || (6:60) | 20:1} 729) SVP oes Ses cee noel toe Geen Gee ete a BSN Soe |) coe] |] eRe) LO Emmett and Davisson.............. 1.66 | 0.67 | 6.61 | 1.19 | 5.70 | 21.4] 8.0 Miversrand Hines = 21:4. sec ciiere shies: 1.68 | 0.70 | 6.34 | 1.32 | 5.98 | 21.9] 8.0 (GrandwAverage.a.ciee acces. a: Dod | Leta loro on L400 ronan | zllednlsnO 1 Omitted from average. 2 Used standard sent out. The triplicate results of each chemist, as a rule, agree very well for the three samples. The summary of the data indicate that there were very marked differences between the laboratories. These may have been due to faulty technic or to possible errors in sampling the products. 276 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 TABLE 4. Percentage deviation from the arithmetic mean. BEEF EXTRACT A BEEF EXTRACT B CER EEAIOLNG ENN fo ee es ee ee el Bomar ANALYST DESICCATED IN 25 cC. OF MEAT . sos . ee Creatin Creatinin | Creatin Creatinin | STANDARD per cent per cent per cent per cent per cent per cent MoH So cebadeoods obs 3.8 50.4 1.3 4.3 3.6 4.6 FOO Ketel ee estrone rs 7.0 4.7 18.5 Pasa 5.4 6.6 RTA MACK ee eers eee 13.5 42.4 10.7 36.1 14.8 5.6 (Sieihh al ely gUbae eee Dae aio.G 1325 6.3 al pions Pra 4.2 Emmett and Davisson.. . 5.7 47.2 10.4 15.0 0.9 0.4 Myers and Fine......... 7.0 44.9 5.8 5.7 4.0 2..8 AVCLABC...052 creme 8.4 32.6 8.9 14.6 5.7 4.0 That this difference in these data is not due entirely to the sampling is evident from the results reported upon the standard creatin solution. Here there seems to be great variation, ranging from 19.9 to 22.3 per cent and averaging 21.3 per cent. Calculating the percentage deviation from the arithmetic mean for these data, Table 4 last column, it varied from 0.4 to 6.5 per cent averaging 4.0 per cent. In other words from these lim- ited data representing the results from six different laboratories it appears that the errors in the Folin method for determining creatin in astandard solution averages about 4 per cent. In order to obtain information regarding the possibility of the samples not being thoroughly representative, Beef extract B was sent out about 23 months after the other samples had been forwarded. The necessity of asking for this further codperative work became evident as soon as the results for the meat and the Beef extract A were reported. In the case of Beef extract A, the directions did not specify that it should be mixed just previous to weighing the portion for analysis. With Beef extract B, the analysts were requested to mix it very thoroughly. This was done by first warming the extract slightly, and then stirring. It should also be stated that Beef extract B was taken from the same lot of beef extract as Beef extract A. The lot was warmed and very carefully mixed in both cases before sampling. In comparing the corresponding data (Table 3) for these two samples of beef extract, it is seen that the results for Beef extract B show less variation between the different laboratories than do those of Beef extract A. This fact is brought out more clearly in Table 4 which gives the per- centage deviation from the mean. The average values are: For creatin 32.6 and 14.6 per cent respectively for A and B, and for creatinin 8.9 and 5.7 per cent respectively. The latter value for creatinin is not far from that found for the standard creatin solution. 1915) EMMETT: SEPARATION OF NITROGENOUS BODIES 277 These data are significant and indicate that the portions that were weighed for analysis from Beef extract B were more representative than were those from Beef extract A. It would seem, therefore, that the varia- tions in the data for A were due in part to a physical change in the sample after it was sent out, that is, crystalline bodies may have separated out to a greater or less extent. Here again it is evident that more data would assist greatly in arriving at a more definite conclusion as to the per cent of errors in the Folin method for determining creatin and creatinin in beef extracts. The data for the creatin in the desiccated meat show less variation than in the beef extract. The sample of meat was naturally easier to mix, and being dried there was no possibility of having any such changes take place as might occur in beef extract. Even these data, however, do not agree as closely as one might expect. The minimum per cent of cre- atin was 1.36 and the maximum was 1.78, giving an average of 1.57 per cent for all the data. The percentage deviation varied from 3.8 to 13.5, averaging 8.4. These results vary about as much as those that were reported last year, as the following data upon desiccated meat show: Percentage of Percentage Analyst creatin deviation IWIRO) VULNS 315 8s arsine Sensi eRe Rea Cees oe ae ce eeseari sine creat 1.80 4.6 (CHT es dal Olah Be Nee ton ee ne oe ie lied EP le een yy Br 1.88 9.3 Eger carats Kereuny AGE AUT OT ere uy -5 cetera, rated calor ieiatoxelevekalens Sierotersis lavatetesiatersie ste 1.50 7.0 BING GETAT COME) A VISSOM; 4.07 3.69 3.71 3.57 4.07 3.67 +0.40 3.83 3.59 | +0.24 9.56 9.17 9.23 9.01 9.67 9.86 | + 9.56 9.25 9.20 9.21 9.60 9.21 +0.39 9.26 9.11 +0.15 average per cent 6.72 6.26 5.45 4.89 4.02 4.31 7.67 8.00 7.63 UNDRIED SAMPLE TABLE 2. Effect of petrolic ethers from different sources on the determination of fat. (9 hours’ extraction; 3 hours’ drying) DRIED SAMPLE 1915] JONES: FEEDS AND FEEDING STUFFS 299 TABLE 2—Continued. UNDRIED SAMPLE DRIED SAMPLE Diff Di nae = | MOISTURE Pennsylva- areas setrolioe Pennsylva- Kansas auorrnce See || pentane || Seen} Bee | ipentane), ||; ener and ee We cance, || Resse ||) ar cemelan | ERS Pea EI Saar cent 9.27 9.01 9.05 | 8.99 9.24 hs 9.03 i 5.98 9.10 8.98 9.05 8.97 9.20 9.00 | +0.20 9.04 8.98 | +0.06 7.61 7.24 { 7.251) 7.16 | 7.58 724 12 4.36 7.66 7.21 puese j 7.22 7.62 7.23 | +0.39 7223) | TA ION NEE OROS 3.32 2.97 2.81|] 2.58 | 3.24 2.7 13 6.98 313|{ 2.88 2-76 2.58 3.23 2.93 | 0.30 2.76 2.58 | +0.18 , 6.20 5.61 | 5.22 5.31 6.17 5.25 14 7.72 6.21 5.66 5.23 5.29 6.19 5.641 0.55 5.23 | 75.30| —0.07 4.67 4.54 3.95 |] 4.30 4.67 3. 15 7.10 477 4.40 4.14 4.29 4.70 4.47 | +0.23 4.01 4.30 | —0.29 5.15 | 4.85 |) 4.84 4.76 5.15 4.8 16 6.70 5.23 4.82 i 4.78 4.77 5.18 4.84 | +0.34 4.84 4.77 | 0.07 8.76 | 8.35 | 8.81 8.31 |) : 8.80 18 7.15 8.88 8.28 8.71 8.04 | 8.81 8.32 | +0.49 8.76 8.18 | 10.58 4.29 | 3.73 3.80 3.50 | = 4.31 -91 | > shhaes 20 6.35 4.21 3.75 3.90 3.63 4.97 3.74| 40.53 3.87 3.57 | +£0.30 9.48 | 9.08 8.76 9.11 | 931 8.78 at 21 5.94 9.30 8.99 8.93 8.96 9.36 9.04 | 0.32 8.82 9.04 | —0.22 3.77 3.61 2.74 2.98 3.80 2.67 2a Gea) 3.78 3.51 i264 2.75 3.78 3.56 | 0.22 2.68 2.87 | —0.19 300 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. J, No. 2 TABLE 2—Continued. UNDRIED SAMPLE DRIED SAMPLE Diff Diff Sal MOISTURE Pennsylva- Kenriser sateolie’ Pennsylva- Kansad petrolic! mis ipetrolicl| "pentane. |) ese enerel| MR PEUOUE) | pentane || Meneres = = BO Oe Pe eee ||| cecaes way) pmmerreentas| “(Co cere | Cecwgareal| \ emerge 3.77 2.05 1.84 1.70 |} . 3.83 |} iy ear 1.74 23 5.64 3.74 1.95 1.68 1.76 || 3.78 2.00 | +1.78 1.75 1.73 | +0.02 2.01 0.94 1.53 0.93 2.03 \ eerie 24, 11.08 1.98 0.90 1.32 0.21 | | 2.01 0.92 | +1.09 1.40 0.92 | +0.48 4.28 4.09 4.01 3.95 3 4.30 1h eatos 25 9.26 4.39 4.04 3.99 4.04 4.30 4.07 | -+0.23 4.07 4.00 | +0.07 present and the increase in extract, hence while it is apparent that the presence of moisture affects the determination the amount of this effect is more largely controlled by the nature of the material extracted than by the amount of moisture present at least when this does not exceed 8 per cent. It is also apparent that the extracting power of petrolic ethers from different sources at least of the two used in this investigation is not affected to the same degree by the moisture present in the sample. OFFICIAL COMPARED WITH PETROLIC ETHER METHOD. Comparing the results with the proposed petrolic ether method (3 hours’ extraction) and the official method all the samples except two, 18 and 24, show an appreciable increase in the amount extracted by the latter ranging from 0.10 per cent for Sample 10 to 1.53 per cent for Sample 1, with an average increase for the 21 samples of 0.62 per cent. The most pronounced differences are shown by the distillers’ and brewers’ grains and compounded feeds containing them, although Sample 25 (wheat bran and screenings) gives a difference of 0.97 per cent. In the case of cotton- seed products appreciable differences are found in that Sample 3, choice cottonseed meal, shows an increase by the official method of 0.49 per cent, Sample 8, good cottonseed meal, 0.64 per cent and Sample 9, cotton- seed feed (meal and hulls), 0.34 per cent. No reason was apparent for the failure of Samples 18 and 24 to follow the general rule of decreased amount of extract with petroleum ether the former showing an increase of 0.04 per cent and the latter 0.16 per cent. 1915] JONES: FEEDS AND FEEDING STUFFS 301 When the petroleum ether extraction is continued for 9 hours the dif- ferences are not so large and in six of the samples 2, 7, 9, 10, 18, and 24 the extraction with petroleum ether has increased so that it exceeds that of absolute ether. Seventeen of the samples show an average excess by the official method of 0.40 per cent ranging from 0.02 per cent for Sample 22 to 1.21 per cent for Sample 1. Six samples show an average excess for petroleum ether of 0.20 per cent ranging from 0.04 per cent for Sample 7 to 0.36 per cent for Sample 24. COMPARISON OF RESULTS OBTAINED BY THE USE OF ABSOLUTE ETHER ON DRIED AND UNDRIED SAMPLES. From Table 3 it is seen that in comparing the official method with the method of extracting the sample with absolute ether, without previously drying, increases in the amount of extract are obtained in the case of 12 of the undried samples. These increases vary from 0.03 per cent in the case of Sample 9 to 1.29 per cent in Sample 21 with an average increase of 0.49 per cent. In 11 of the samples slightly higher results were secured by the official method on the dried samples, the increases varying from 0.01 to 0.29 per cent with an average increase of 0.13 per cent. In Samples 1 and 21, which are distillery by-products the increase of extract obtained from the undried samples is very marked, being 1.11 and 1.29 per cent, respectively, while with the other samples the range is from 0.03 to 0.60 showing that the nature of the material under examination affects the quantity of extract secured from the undried condition as compared with that obtained by the official method. In comparing the relative value of different feeding materials it is es- sential that a moisture-free basis should be secured and it seems advisable from the results of investigation to continue the extraction of the crude fat from the moisture-free substances. This is especially necessary in order to compare the work of different analysts on the same basis. EXTRACTION WITH SQUIBBS ETHER AS COMPARED WITH OFFICIAL METHOD. In view of the fact that the official method requires the use of ethyl ether free from water and alcohol, it is concluded that the presence of these constituents in the solvent exerts an influence which materially affects the quantity of ether extract. Therefore, it was deemed advisable to study the effect of using Squibbs ether containing approximately 3 per cent of alcohol as compared with absolute ether. In 16 of the samples analyzed the increase in extract obtained with Squibbs ether varied from 0.03 to 2.99 with an average increase of 0.57 per cent. In one sample no change was observed. In six of the samples the results showed a decrease in the quantity of extract obtained with ether containing alcohol, varying from 0.02 to 0.21 with an average of 0.12 per cent. sooo if | OR aes I 92:0+ | 06'T soo f 1 rie WN bP I 90'0+ | €8'0T een KOLWOTa \ | L801 f 80°0+ | 28°9 soo | OHO I \ | wo fs 66° S+ | 24ST sosodl |) OMe) \ | ersrs quao sad Meee, - RAteeee e[ dues -woosayje| PMA sqqinby (DNIAUC SUNOH FT ‘NOIMOVULX ,SXN0H QT) UGHLG saainds quad sad = TetoyO IIA persed -u100 ‘But -AIp ,s1noy § ‘dorzoBI4 -x9 SInoy § 1eq}0 uINd[O1}9g s9'0— | 62:0— 69'e | 08's 10'0+ | 08% 18% 92'0+ | 89°% ie soa Tsao | ueseoate leery (e leeeet cuelitga' cel elpereeel| ee ier | Grete \ ee |) O26 Ti . 61° Le'0—_| ¥0'0— me | One 40'0+ | 8¢°0 | s9°0 9t' 0+ 10 See | eer oer ee Ieee ero ae \ 60 T | FOT fs \ sso 6| e9'°0 f 6P'0 91:0— | 80°0— 1L°01_| 69°0T St o+ 98°01 | 19°0T st 0+ eh OT eae: evi RDsOTMEILEO OTe lense lar |e OlM|EcoFOIN A \ LL 01 | 69°01 f \ cr OL | 68°01 J rr OL 91'0+ | 11'0— 629 | 8I'9 ¢0'0+ 60°9. | 219 czo+ | 029 =e Bi eee sea | i wire ieee. 1 ae \ 2G) | 4see) yf \ 809 | 009 f 62'9 12@:T= | 11+ Sh'ZI_| 69° ST €0°0+ SL IL | St‘ It 1h 0+ 98° OT poe, Puen lpers7cta leyLaeye eee le se OTe Ties Geum) ea \ eh cl | PP SI J \ PL'IL | St Its 90° TT muon sad | quorvad | PG | yheoton | 4d | seooiea | quoo moa | M™ewed | iol log (ow qt peed puss ‘But = = = ~Al sino & heen Lae Por 9] dures ee a[ dures ojdures AWK a] dures -xo sinoy peeiias e[duies pelupug saLhiecs peuqdg pempug permduzoo pod Moca a eePaceurie alee a pexpug peupun e|dures paipug DNIAUG §, 40H § ‘NOILOVYLXG ,SUN0H Q[ ‘UAHLA aLAIOSav (ONIAUG ,SUNOH F !NOILOVULXG ,SHNOH 6) ANVINGd yof fo uoyvurusajap ay, uo aumgswow fo yal ‘€ TIGVL qo0442a quao sad e[dures peupuy (DNIAUG ,SHOOH g !NOILOVHLXA {SU00H 6) AGHLT O1louLaa 68'P Sp ¢ 929 cL 9 quao tad abpusap GuoLsion 302 “ON a1dNvs ez0+ | oar gzI— | 62:0— | 1r0— | zo | t6'¢ ceo+ | gc:2_ | g6'z bot | one | eae te. Oger |e oa eee oleh Whaine ise 1 Wee sie 1 Die IPEIES 86°9 0% «| Fe { 6a'F f \ 80% | zor J \ soe | 26% J iz |zee | 9ro+ | e8'2 er0— | so0— | 0'0— | 292 | e9:2 70'0+ | 6r'z | gaz peo+ | saz | coz Soil UR Lee | eee Ee NED LEA anced pees VA ERD V Scn 984 | 992 98°F Hail | Soe POR 2 I l 69'2 | 692 f \ ee Nie If en | 192 zo'0 — | 0F'6 I¢'0— | zz0- | 100- | 26 |1%6 | 200+ | 968 | 00'6 910+ | 40°6 | 02'6 SPA ClhOR Ginette WN Sea Weel are Wee Wl aca | ae 96'S ; £0°6 | #26 \ | 0F 6 \ Sh '6 oF'6 f \ 66°8 10°6 f ¢0°6 126 | 200+ | ze'6 or'o— | te'0+ | ¢c:0+ | 62:6 | z¢°6 or'o+ | 11:6 | 12°6 re0+ | 9276 | 09'6 en mle eae eee i | eecy Neg ) fN eG Were Wl acne OGG) 98" 6 89°L 9¢°6 | 29'6 \ | 086 \ 66 | ee6 f \ 106 | 216 f 9z'6 | 99°6 or'o— | 29°¢ re0- | orot+ |e0'0+ | 26'¢ | 00% 300+ | 6g'e | 29°¢ $e 0+ | ese | 20°F Pepin E6826 eli earee =| ener [AN 2ere |f00rn. | Terra WCE When d The | 20°F 00°8 C6 . zee | 2 UU are i \ wor | 668 Jf \ o9'e | to's f 16°& | 20°F “Tro+ | g0'z 79'0— | ce'0— | 90'0— | #12 | 8072 ez'0+ | zo'9 | 22°9 11:0-+ | 2979 | 62:9 (ite A008 a | \ceeeee. Peeeee {peek lee 1 ff BHO 1 GPO Wl anor 879 | 999 19° L FO'L e¢9 | €8°9 eI Aor fp l OP arr I \ 199 | 129 f 9¢'9 | 89°9 10°0—_| #9°9 ez'0—- | 400+ | 60 0- | 12:9 | zo'9 co'0+ | 98:9 | 1F'9 sto | 269 | 9279 { }Ob9 Y] | one { | e909 | ¢9°9 | Mee Ope lees, ol ener eeu OR) Seca eS |e8ocony) \ 919 199 J \ re°9 1g9 J cog | 22'9 1Z'0— | oz'¢ os'0— | oc o— | ez 0—- | te | ate 70'0—- | 28% | e8°2 120+ | 02% | 16% soy (ZS [Eaponag™ |Pesccos Joa| gece | o%e a Sf Wereree Wenice| eee | areae 18°F lee Eot- Be aa ae \ | 2 | 96% J iuilbeereeealingaes 303 09°0+ | 02'¢ 89'0— | se'0— | sr'0+ | 09% | g0'¢ 40+ | s¢7e | P2'e oro+ | zee | 247 { re} Sele an eA Pele | eeoig) geen et mes ve oe ce'9 | 0g Ga \ oor | eros \ OGRemea ne zney) oss | 6'F 6L0— | 28°8 FO'0O+ | 620+ | 227:0+ | 9¢°8 | ¢8°8 meas |) es ae" co'0+ | 92'8 | 188 SRE ee leo, NEST | CN eas ee WE oceans Wa ov | st LPS J \ ESAS) Kez eh) ll [Sy 1) GA Ih \ 18°8 || gy-g 0z'0+ | ¢¢'¢ Ir0— | 2400— | 2zo+ | se¢ | szg'¢ Lo;0+ | 22h | t8'°F re 0+ | 8h | Sto GcocnN ee ell Peer sei | Seep ROO; seer. aye caer Bees alecna tp IeOan pen reo J \ Teg | egos \ OL | G8 hs say Gre 00'0 | 08’ Te'0— | or 0— | 820+ | 08% | go's 10+ | 08% | 2r'F 69° 0+ 10% | OL F GO eae le |e eee, ea eee raf qg'y | 119) mul Tei WGP 1) eos ae fae Ol'L CT C8 Ff \ GLY Go's J \ 08°F oF) ce 19°F 90°0+ | #9°9 69'0— | 6z:0— | 090+ | SF 9 | 8072 Fe'o+ | ogg | F9°¢ 96°0+ | g2'¢ | 61:9 PE CEPON Mesias f | WO] oo, Fife 16ZaGe wiROGAoNe alt ee Gg | Te'9 zk | PI ' ae 80°2 ; cog =| 209 ee RESEON \ o9'9 f \ Tes | 19°Sf zz¢ | 0z°9 muon sad | MooNeg | mended | guooued | qusoved | Tore | groin | meow | Fo | coowed | tueowed | 4% | suo ted | “pousoun 1eIOWO [e1oWO q}IM posed | 441 poied = Erp anon | | wot | ore | 2ep | zo+ |_ eet 00% |eoe+ | ont | gute ee ’ {| se [geet | oer |e... Seas 19's \ PSE | str) OLT | 90's) ast | ze F10— | co'0— | se'ot+ | ose | sty | eo'0+ | wz loose | orrt+ | soz | sz'¢ cae Olea aber solk gi BO24G- (/cOTsP aa ea ce ple ter acaea ara Piles eres ence 6I'6 \ | we | 0cF \ | 86% | 19°8f wz «| Le { 62 I= | 160- | 61+ |_zeor | 9c'1r | o00+ | 6 | 06 |9:0+ | zg | 86 nee lane Pee Sct cmnciilradealngre teste ee ecaealnecee 76'9 L_ | Seon | 2911s NIG ea eo Key) 92:8 | 886 305 306 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 It is very evident from the results obtained, not only in this investiga- tion but also in comparative work in connection with inspection of feeding stuffs of varied composition in this laboratory that the use of ether con- taining alcohol and water can not be relied upon to give absolute results. The nature of the material subjected to examination appreciably influences the solvent action of the ether as is shown in Table 3. For instance, Sample 1 composed chiefly of corn distillers’ grains and Sample 21, which is corn distillers’ grains with no admixture, gave an increase in amount of extract with Squibbs ether of 2.99 and 2.78 per cent, respectively. With gluten feeds increases were apparent, but with linseed and cotton- seed meals either slight increases or negative results were secured. Con- sidering the widely varying results obtained with Squibbs ether as com- pared with absolute ether, due not only to the presence of alcohol, but also to the nature of the material under examination we would emphasize the importance of using absolute ether by all analysts in order to secure results which are comparable. EFFECT OF TIME OF DRYING EXTRACT. Repeated investigations in this laboratory having shown that drying ether-extracted residues 13 hours at the temperature of boiling water gave constant weight, it was deemed advisable to determine the length of time necessary to dry extracts with petroleum ether to secure similar results. Two additional periods, multiples of 13, 3, and 43 hours were decided upon. Nineteen of the undried samples with Pennsylvania petrolic ether show an average loss of 0.12 per cent from 14 to 3 hours, while 4 show an average increase of 0.09 per cent with a general loss for the 23 samples of 0.09 per cent. For the 3 to 43 hour period 14 of the samples show an average in- crease of 0.07 per cent and 7 a decrease of 0.04 per cent with an average increase for the 23 samples of 0.03 per cent. The dried samples show a similar variation in that all 23 samples show an average decrease from 14 to 3 hours of 0.19 per cent, while from 3 to 41 hours 15 show an average increase of 0.09 per cent, while 7 show a decrease of 0.03, 2 show no variation, making the average increase for 23 samples 0.05 per cent. In the pentane extraction 21 of the samples show an average loss from 134 to 3 hours of 0.25 per cent, while 2 show an average gain of 0.07 per cent, making the average loss for the 23 samples 0.22 per cent. Thirteen show an average loss of 0.06 per cent from 3 to 44 hours and 10 a gain of 0.10 per cent. Twenty of the dried samples show an average loss of 0.14 per cent and 3 a gain of 0.07 per cent from 13 to 3 hours, making the average net loss 0.12 per cent, while from 3 to 44 hours 14 samples con- 1915] JONES: FEEDS AND FEEDING STUFFS 307 tinue to lose an average of 0.07 per cent and 9 gain 0.11 per cent. The per cent of loss in all samples, however, is such as to indicate that prac- tical work will require not to exceed 3 hours’ drying of the extract. In all cases of drying the extract the length of time required would be materially affected by the completeness of the drying before the flask was removed from the extraction machine. It will be noted that much greater losses were shown in last year’s report between 2 and 4 hours. DISCUSSION. The difficulty of volatilizing the petroleum ether noted in the report of 1912 was again experienced, and in no case were the analysts able to distill the entire residual petroleum ether into the Soxhlets; it was necessary to volatilize this out of the receiving flask into the open. Strict account was kept of the amount of Pennsylvania petrolic ether and absolute ether used with a view to calculating the cost per determination with each method. It was found that 51.9 per cent of the petrolic ether was lost and that for 24 determinations the total loss amounted to one pound, which on the basis of October 17 quotations would cost 35 cents or 1.5 cents for each determination for reagent. In the official method the loss in ether was 40 per cent, the total loss for 24 determinations, amounting to 0.8 pound, which on the basis of October 17 quotations from stock (80 cents a pound) amounts to 64 cents or 2.7 cents for each determination, and on duty-free import (373 cents per pound) to 30 cents or 1.3 cents for each determination. CONCLUSIONS. The work of the past three years as well as past work of the association’s referees leads to the following conclusions: (1) With very rare exceptions appreciably higher amounts of crude fat may be expected on all classes of feeding stuffs from the use of the official method than from the use of the proposed petroleum ether method. (2) That 3 hours’ extraction with petroleum ether is not sufficient to secure the total petrolic ether extract, which investigation indicates re- quires 9 hours’ extraction to secure maximum results. This is true even with cottonseed products for which the method is recommended by the previous referee. (8) That the presence of moisture materially affects the amount of ex- tract by the petroleum ether method depending on the nature of the material extracted. (4) That the extraction of the same materials with petrolic ethers from different sources does not give concordant results, that is, petrolic ethers ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 308 TABLE 4. BAMPLE | vorsr URE ways per cent 1 6.72 2 | 6.38 3 5.45 4 4.86 5 3.98 6 4.27 7 | 7.58 8 | 7.67 9 8.00 10 7.63 Effect of time of drying extract. (9 hours’ PETROLIC ETHER Undried sample Dried sample Gain or | Gain or Gain or 14 hours’ | 3 hours’ | 44 hours’ |loss 14 to 3/loss 3 to 43} 1} hours’ | 3 hours’ | 43 hours’ |loss 14 to 3 drying drying drying hours’ hours’ drying drying drying hours’ drying drying drying percent | percent | me. cert | per cont | percent | Be,cont | pez | oe cent Saree TT S16 11.35 | +0.05 | +0.19 11.08 11.06 11.12 | —0.02 11.08) 11.26 11.38 | +0.18 | +0.12 10.94 10.81 10.97 | —0.138 11.31) 11.38 11.52 | +0.07 | +0.14 10.94 | 10.72 11.26 | —0.22 LUGL7| N27, 11.42 | +0.10 ) +0.15 10.99 10.86 115125 =0st3 6.35} 6.50 6.55 | +0.15 | +0.05 6.41 6.29 6.38 | —0.12 6.29] 6.45 6.49 | +0.16 | +0.04 6.18 6.28 6.28 | +0.10 6.38} 6.41 6.52 | +0.03 | +0.11 6.18 6.04 6.08 | —0.14 (6.34, 6.45 6.52 | -+0.11 | -+0.07 6.26 6.20 6.24 | —0.06 Sausieis 10.48 10.54 sone || seOils syed 10.44 10.36 tee 10.76) 10.67 10.68 | —0.09 | +0.01 dias 10.39 10.30 oe 10.70) 10.68 10.71 | —0.02 | +0.03 10.53 10.45 10.41 | —0.08 10.73) 10.61 10.64 | —0.06 | +0.03 10.53 10.43 10.36 | —0.08 0.80} 0.68 0.76 | —0.12 | +0.08 0.73 0.49 0.49 | —0.24 0.75} 0.66 0.72 | —0.09 ||| =-0.06 0.54 0.52 0.49 | —0.02 0.98) 0.68 0.71 | —0.30 | +0.03 0.60 0.53 0.54 | —0.07 0.84) 0.67 0.72 | —0.17 | +0.05 0.62 0.51 0250)/ Ont 2.95 2.91 3.07 | —0.04 | +0.16 3.06 2.79 2.80 | —0.27 3.05 2.99 3.07 | —0.06 | +0.08 2.84 2.56 2.66 | —0.28 2.99 2.91 3.01 | —0.08 | +0.10 2.96 2.01 2.76 | —0.25 3.00} 2.94 3.05 | —0.06 | +0.11 2195 2.69 2.74 | =0226 3.03 2.90 320370213) || (OMS 2.94 2.71 2.76 | —0.23 2.97 2.87 3.02 | —0.10 | +0.15 2.91 2.68 2.68 | —0.23 3.10) 2.96 Sly || SOAS |] SAD sih! 2.90 2.70 2.70 | —0:20 3.03) 2.91 3.04 | —0.12 | +0.13 2.92 2.7 2.71 | —0.22 6.7. 6.77 6.87 | +0.04 | +0.10 6.65 6.55 6.48 | —0.10 6.84 6.79 6.86 | —0.05 | +0.07 6.54 6.60 6.51 | +0.06 6.61 6.68 6.78 | +0.07 | +0.10 6.54 6.57 6.54 | +0.03 6.73 6.75 6.84 | +0.02 | +0.09 6.58 6.57 ee) Soi 6.97; 6.88 6.98 | —0.09 | +0.10 6.84 6.56 6.54 | —0.28 6.91 6.83 6.89 | —0.08 | +0.06 6.72 6.53 6.54 | —0.19 6.75) 6.66 6.76 | —0.09 | +0.10 6.68 6.48 6.46 | —0.20 \ 6.88} 6.79 6.88 | —0.09 | +0.09 6.75 6.52 6.51 | —0.23 4.18) 4.02 3.94 | —0.16 | —0.08 4.19 3.97 4.08 | —0.22 4.19) 4.12 4.08 | —0.07 | —0.04 4.04 3.82 3.91 | —0.22 4.16} 4.07 3.98 | —0.09 | —0.09 3.97 3.71 3.76 | —0.26 4.18 4.07 4.00 | —0.11 | —0.07 4.07 3.83 3.92 | —0.24 9.50) 9.56 9.51 | +0.06 | —0.05 9.35 9.23 9.31 | —0.12 9.51 9.67 9.59 | +0.16 | —0.08 9.51 9.36 9.43 | —0.15 9.36) 9.56 9.50 | +0.20 | —0.06 a 9.20 9.21 ie 9.46} 9.60 9.53 | +0.14 | —0.07 9.43 9.26 9.32 | —0.13 1915] JONES: FEEDS AND FEEDING STUFFS extraction at temperature of boiling water, 98°C.) PE- TROLIC PENTANE 309 ETHER et Undried sample Dried sample Gan Gain or Gain or Gain or Gain or Os vr 1} hours’ | 3 hours’ | 44 hours’ |loss 14 to 3}loss 3 to 44] 14 hours’ | 3 hours’ | 44 hours’ |loss 14 to 3/loss 3 to 44 3/0 H drying drying drying hours’ hours’ drying drying drying hours’ hours’ dou drying drying drying drying drying : 3 | ORB ol ete per cent|) ewe | “actract | aetract | Percent | percent | Recent | Decent | mer cent || per cent'| per cent ae 138) i145 | 11-33-| —0.23 | +0.18 | 11.28 | 11.14) 11.12) =0.14| =0.02 Fo'sa| [11-46] 11.20) 11.23 | -0.26 | +0.03] 11-31] 11-16] 11.21 | —0.15 | +0.05 +0.26} 11.42| 11.18 | 11.28 | —0.24|+0.10| 11.30| 11.15 | 11.17 | —0.15 | 40.02 +0.09| | g 241 6.00} 5.97| —0.24| -0.03| 6.00] 6.08| 5.96| +0.08 | —0.12 +004] | 8-45] 6.23] 623] -021] 0.00] 6.05] 6.10] 6.01 | +0.05 | —0.09 +004) 6.35]. 6.12| 6.10 | —0.23| —0.02| 6.03| 6.09 | 5.99 | 0.06 | —0.10 oe 10.65] 10.39 | 10.39| —0.26| 0.00] 10.46| 10.42] 10.16] —0.04 | —0.26 =) -b4| {10.84 10.62] 10.60 | —0.22/ —0.02] 10.43] 10.30] 10.18 | —0.13 | —0.12 =0.07) 10.75| 10.51 | 10.50 | —0.24|—0.01| 10.45 | 10.36| 10.17 | —0.09 | —0.19 0 pel | 0.73} 0.63] 0.59| —0.10] —0.04| 0.71] 0.58] 0.52 | —0.13 | —0.06 xo o1| | 0-83] 0.67) 0.66] -0.16|-0.01] 0.70) 0.57] 0.44] —0.13 | —0.13 =—oilmmds7s| 0165 | 0:63 | —0.13 | =0.02| O.71| 0.58 | 0.48 | —0.13 | 20.10 ae 1G 2.93} 2.76| 2.74| —0.17| —0.02| 2.93] 2. 2.81 | —0.10 | —0.02 o.05| | 2-95] 2-86] 2.85 | —0.09| —0.01] 2.85] 2.77] 2.67 | —0.08 | —0.10 +0.05 2.94, 2.81| 2.80| —0.13|—0.01| 2.89| 2.80| 2.74|—0.09 | —0.06 ee 3.14 2.95| 2.94] —0.19| —0.01| 2.84] 2.79| 2.70] —0.05 | —0.09 ooo | 2-87] 2.70| 2.66) —0.17| —0.04] 3.00] 2:94] 2.85 | —0.06 | —0.09 +0.01| 3.01) 2.82| 2.80|—0.18|—0.02| 2.92| 2.87| 2.78 | —0.05 | —0.09 | yao 07) | 6.41] 6.37| 6.43] —0.0¢|+0.06| 6.57] 6.34] 6.46 | —0.23 | +0.12 ch us| { 6.42) 6.45] 6.54] +0.03| +0.09]| 6.50] 6.38] 6.53] —0.12| +015 706) 6.42| 6.41 | 6.49 | —0.01| +0.08| 6.64] 6.36| 6.50| —0.18 | +0.14 Rec: 6.92 6.71| 6.70 | —0.21| —0.01| 6.81| 6.51| 6.56] —0.30 | +0.05 Don] { 6.98} 6.82] 6.81] —0.16|—0.01| 6.81} 6.53| 6.57) —0.28 | +0.04 Holl) 9 6.95| 6.771 6.76 | —0.18|—0.01| 6.81| 6.62| 6.57 | —0.29 | -£0.05 Be cg 3.68} 3.641 3.49| —0.04|-0.15| 3.86| 3.60| 3.65] —0.26| +0.05 Foos| | 3-89| 3.69] 3.71] -020|+0.02| 3.77) 3.57] 3.61 | —0.20 | +0.04 +0.09| 3.79] 3.67 60 | —0.12 | —0.07 32 | 3.59] 3.63 | —0.23 | 40.04 +0.08' | 934, 9.17] 9.37| +0.03| +020] 9.14] 9.01! 9.37] —0.13 | +0.36 +0.07 Fooi| | 9-09] 9.25] 9-43] +0.16) +018] 9.14] 9.21] 9:51 | +0.07 | +030 +0.06} 9.12. 9.21 | 9.40 | +0.09 | +0.19 14| 9.11| 9.44 | —0.03 | 40.33 310 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMIsTS [Vol. I, No. 2 TABLE 4.—Continued. PETROLIC ETHER Undried sampl Dried sampl BAMEUE MOISTURE: narie mple rie mple Gain or | Gain or Gain or 1} hours’ | 3 hours’ | 44 hours’ |loss 14 to 3/loss 3 to 44] 1} hours’ | 3 hours’ | 43 hours’ |loss 14 to 8 drying drying drying hours’ hours’ drying drying drying hours’ drying drying drying per cent | Ravcent | percent | percent | percent | percent | Beycent | percent | percent | per cen 9.43} 9.27 9.27 | —0.16 0.00 9.48 9.05 9.14 | —0.43 ml 5 98 9.22) 9.24 9.23 | +0.02 | —0.01 ee 9.03 9.03 yee : pone 9.10 9.17 ere ote OnOu Meeks 9.05 9.09 S608 9.33) 9.20 9.22 | —0.07 | +0.03 9.48 9.04 9.09 | —0.43 8.08) 7.61 7.68 | —0.47 | +0.07 ae 7.25 7.25 12 4.36 7.66| 7.58 7.64 | —0.08 | +0.06 ae 7.24 7.33 Sooo ; 7.69 7.66 7.68 | —0.03 | +0.02 7.46 7.36 7.55 | —0.10 7.81 7.62 7.67 | —0.19 |} +0.05 7.46 7.28 7.38 | —0.10 3.60} 3.32 3.33 | —0.28 | +0.01 3.09 2.81 2.81 | —0.28 13 6.98 3.48) 3.24 3.24 | —0.24 0.00 3.01 2.70 2.75 | —0.31 : 3.46} 3.13 3.13 | —0.33 0.00 3.04 2.76 2.84 | —0.28 3.51 3.23 3.23 | —0.28 3.05 2.76 2.80 | —0.29 6.44] 6.20 6.23 | —0.24 | +0.03 5.43 5.22 5.19 | —0.21 14 7-72 6.36} 6.17 6.13 | —0.19 | —0.04 5.52 5.25 5.27 | —0.27 ; 6.48 6.21 6.30 | —0.22 | +0.09 5.50 5.23 5.23 | —0.27 6.41 6.19 6.22 | —0.22 | +0.03 5.48 5.23 5.23 | —0.25 5.04) 4.67 4.65 | —0.37 | —0.02 4.11 3.95 3.95 | —0.16 15 710 4.73) 4.67 4.70 | —0.06 | +0.03 4.22 3.93 3.91 | —0.29 : 4.76) 4.77 4.70 | +0.01 | —0.07 4.18 4.14 4.12 | —0.04 4.84; 4.70 4.68 | —0.14 | —0.02 4.17 4.01 3.99 | —0.16 5.33} 5.15 5.18 | —0.18 | +0.03 5.08 4.84 4.79 | —0.24 16 6.70 5.28) 5.15 5.18 | —0.13 | +0.03 4.95 4.89 4.83 | —0.06 ; 5.31 5.23 5.18 | —0.08 | —0.05 4.89 4.78 4.72) | =O.11 5.31 5.18 5.18 | —0.13 4.97 4.84 4.78 | —0.13 33 9] 88] -88] +28] lem] an] aaron 18 7.15 i ; “na ; 8.84 8.71 8.82 | —0.13 8.91 8.88 8.83 | —0.03 | —0.05 8.89} 8.81 8.83 | —0.08 | +0.02 8.90 8.76 8.90 | —0.14 4.47 4.29 4.32 | —0.18 | +0.03 3.91 3.80 4.04 | —0.11 20 6.35 4.40} 4.31 4.40 | —0.09 | +0.09 4.01 3.91 4.15 | —0.10 ; 4.28) 4.21 4.22 | —0.07 | +0.01 4.00 3.90 4.04 | —0.10 4.38] 4.27 4.31 | —0.11 | +0.04 3.97 3.87 4.08 | —0.10 9.57} 9.48 9.56 | —0.09 | +0.08 8.98 8.76 8.79 | —0.22 21 5.94 9.41 9.31 9.50 | —0.10 | +0.19 9.00 8.78 8.75 | —0.22 ; 9.35} 9.30 9.47 | —0.05 | +0.17 9.08 8.93 8.88 | —0.15 9.44, 9.36 9.51 | —0.08 | +0.15 9.02 8.82 8.81 | —0.20 3.86) 3.77 3.77 | —0.09 0.00 3.24 2.74 2.84 | —0.50 29 9.19 3.88} 3.80 3.80 | —0.08 0.00 2.90 2.67 2.73 | —0.23 : 3.84! 3.78 3.75 | —0.06 | —0.03 2.91 2.64 2.70 | —0.27 3.86) 3.78 3.77 | —0.08 | —0.01 3.02 2.68 2.76 | —0.34 1915] JONES: FEEDS AND FEEDING STUFFS 311 PE- TROLIC PENTANE ETHER ee Undried sample Dried sample eae Gain or | Gain or Gainor | Gainor an vr 1} hours’ | 3 hours’ | 43 hours’ |loss 1} to 3}loss 3 to 43) 13 hours’ | 3 hours’ | 43 hours’ {loss 1} to 3}loss 3 to 44 ee 2) drying drying drying hours’ hours’ drying drying drying hours’ hours’ aaiee drying drying drying drying per cent per cent per cent per cent per cent per cent MEP ES extract extract extract DER Cette HEP GR: extract extract extract Bencere ETI CEIEY +0.09) | 9 19/ 9.01} 9.02} —0.18|+0.01| 9.25] 8.99] 9.02] —0.26 | +0.03 -... | £0.06 | 9.24] 8.97] 9.01 | 0.27 | 40.04 +0.05 9719) 9.00 9.03 | —0.18 | +0.03 9.25 8.98 9.02 | —0.27 | +0.04 ee 7.42} 7.24| 7.29| —0.18|+0.05| 7.36] 7.16| 7.24] —0.20 | +0.08 tote | 7-28] 7-21] 7.25] —0.07|+0.04] 7:31] 7.22] 7.39] -0.09| +0.17 0.10. 7.25) 7.23 | 7.27| —0.12 | +0.04| 7.34| 7.19| 7.32 | —0.15 | 40.13 Bait 3.08} 2.97| 2.98] —0.11|+0.01| 3.10] 2.58] 2.71 | —0.52| +013 tp ce} | 3-00] 3-88) 2.80| 012) -0.08| 314] 2.58] 2:76 | —0.56 | +0.18 +0.04| 3.04| 2.93 | 2.89 | —0.11 | —0.04| 3.12| 2.58 | 2.74| —0.84| +0.16 a Gate erstet (5 84) |l=01n71| 50593 |) 545) 5.31 5 22)| 20.14 I =0009 oor, { 6.10 5.66) 5.80) 0.44] +014] 5.39] 5.29) 5.27 | —0.10 | —0.02 6.14| 5.64| 5.82 | —0.50|+0.18| 5.42| 5.30 25 | —0.12 | —0.05 0] | 4.78} 4.54] 4.66| -0.24) +012] 4.42] 4.30] 4.21 | —0.12| —0.09 pp] [ 4.58] 4.40] 4.46|-0.18|+0.06| 4.37] 4.29] 4.27 | —0.08 | —0.02 =0.02| 4.68| 4.47| 4.56 | —0.21|+0.09| 4.40] 4.30| 4.24|—0.10 | —0.06 an 5.09} 4.85| 4.97| —0.24] +4012] 4.5| 4.76] 4.74] —0.09 | —0.02 5.12) 4.82) 498|—-0.30]+0.16| 4.82| 4.77| 4.76 | —0.05 | —0.01 — —0.06 5.11] 4.84 4.98 | —0.27 | +0.14 4.84 4.77 4.75 | —0.07 | —0.02 & 0.16 8.64 8.35| s.46|—0.29]+0.11| 8.19] 831| 8.31]|+0.12| 0.00 40.11) 8.57/ 8.28) 836] —0.29|+0.08| 7.99] 8.04] 8.14 | +0.05 | +0.10 0.14) 8.61| 8.32| 8.41 |—0.29|+0.09| 8.09| 8.18| 8.23 | +0.09 | +0.05 Bp 1 419] 3.73] 3.64| -0.46|-0.09] 3.59] 3.50] 3.43| —0.09 | —0.07 ee | 414 3.75| 3.63| —0.39|—0.12| 3.70| 3.63| 3.54| —0.07| —0.09 +0.21 4.17) 3.74 3.64 | —0.43 | —0.10 3.65 3.57 3.49 | —0.08 | —0.08 +0.03/ | 9.53| 9.08| 9.06] —0.45|—0.02| 9.07] 911! 9.11 | 40.04] 0.00 9.38; 8.99] 8.97|—039|—0.02| 891] 8.96] 8.90] +0.05 | —0.06 —0.01 9.46} 9.04 9.02 | —0.42 | —0.02 8.99 9.04 9.01 | +0.05 | —0.03 Fe og 3.94] 3.61| 3.38] —0.33| —0.23] 3.01| 2.98| 2.95 | —0.03 | —0.03 Foe] | 3-91] 3.51) 3.38|-0-40|-013| 2.79) 275| 2.71 | 0.04 | 0.04 +0.08| 3.93| 3.56| 3.38 | —0.37|—0.18| 2.90| 2.87 | 2.83 | —0.03 | —0.04 312 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 TABLE 4—Continued. PETROLIC ETHER Undried sample Dried ss 1 ee ouheae rie ried sample Gain or Gain or Gain or 14 hours’ | 3 hours’ | 44 hours’ |loss 1} to 3/loss 3 to 44] 14 hour’s| 3 hours’ | 43 hours’ |loss 1} to 3 drying drying drying hours’ hours’ drying drying drying hours’ drying drying drying per cont | REE | ae | Ee! per cant | mer cant | 7G | recent | eee 3.92) 3.77 3.75 | —0.15 | —0.02 2.08 1.84 1.86 | —0.24 23 5.64 3.96] 3.83 3.81 | —0.13 | —0.02 1.96 1.74 1.75 | —0,22 : 3.88 3.74 3.72 | —0.14 | —0.02 1.91 1.68 1.66 | —0.23 3.92) 3.78 3.76 | —0.14 | —0.02 1.98 1.75 1.76 | —0.23 2.14) 2.01 2.00 | —0.13 | —0.01 1.83 1.53 1.63 | —0.30 24 11.03 2.18} 2.03 2.01 | —0.15 | —0.02 1.59 1.34 1.44 | —0.25 ‘ 2.10 1.98 1.97 | —0.12 | —0.01 1.51 1.32 1.42 |} —0.19 2.14 2.01| 1.99 | —0.13|—0.02| 1.64]; 1.40| 1.50 | —0.24 4.23) 4.28 4.21 | +0.05 | —0.07 4.31 4.01 4.13 | —0.30 25 9.26 4.47} 4.30 4.22 | —0.17 | —0.08 4.63 4.22 4.34 | —0.41 7 : 4.34) 4.32 4.28 | —0.02 | —0.04 4.14 3.99 4.03 | —0.15 4. 4.30| 4.24|—0.05|—0.06| 4.36| 4.07| 4.17 | —0.29 from different crude petroleums do not give the same percentage of extract. (5) That the extracts with petrolic ether must be dried at least 3 hours to secure constant weight. (6) That in general the use of Squibbs ether may be expected to give higher results than the official method, although past experience well establishes the fact that this increase is not due to the presence of true elycerids of the fatty acids. (7) That the extraction with absolute ether of samples without previous drying will in general give higher results than the official method. Tn view of the results set forth in Table 3 the referee wishes to emphasize the absolute necessity of all chemists following the official method in every particular if disagreements are to be avoided. RECOMMENDATIONS. It is recommended that the recommendation of the referee in 1911 “That the association recognize the petroleum ether method for deter- mining fat in cottonseed products’? be not adopted for the following reasons: (1) I do not believe the association should commit itself to the policy of recognizing special methods for individual products unless such products can not be analyzed by official methods already in use. Cottonseed 1915] JONES: FEEDS AND FEEDING STUFFS 313 PE- TROLIC PENTANE ETHER Sea Undried sam ple Dried sample Gan Gain or | Gain or Gain or Gain or Baas 1} hours’ | 3 hours’ | 4} hours’ |loss 14 to 3/loss 3 so 43) 1} hours’| 3 hours’ | 43 hours’ |loss 13 to 3/loss 3 to 44 Vieneet drying drying drying hours’ hours’ drying drying drying hours’ hours’ Bice drying, drying drying drying per cont) Bercent | vercent | nercent | percent | percent | Heyer’ | tercent | percent | er cent | per cen Biya 2-48] 2,05] 1.92 —0.43| —0.13| 1.82] 1.70] 1.62 | —0.12 | —0.08 0.02 | 2.39 1.95 1.83 | —0.44 | —0.12 1.84 1.76 1.69 | —0.08 | —0.07 +0.01 2.44 2.00 1.88 | —0.44 | —0.12 1.83 1273 1°66) |) —0F 10") 0207 Hi ia ] 1.28] 0.94] 0.80| -0.34| —0.14| 0.97| 0.93| 0.87 |=go¢| —0.06 +010 | 27 0.90 0:79 | —0.37 | —0.11 0.96 0.91 0.83 | —9.95 | —0.08 +0.10 1.28) 0.92 0.80 | —0.36 | —0.12 0.97 0.92 0.85 | —0.05 | —0.07 to io] | 4-03} 4.09] 4.14) +0.06] +0.05] 3.98] 3.95] 3.92 | —0.03 | ——0.03 +0.04 4.01 4.04 4.02 | +0.03 | —0.02 4.10 4.04 3.99 | —0.06 | —0.05 +0.10 4.02} 4.07 4.08 | +0.05 | +0.01 4.04 4.00 3.96 | —0.04 | —0.04 products are not of such a nature as to preclude determination of the crude fat by the present official method. (2) The time of extraction proposed for the petroleum ether method does not give maximum results and when the extraction and time of dry- ing the extract is extended to accomplish this result, very little if any time is saved. In addition the question of time and cost should not be the controlling factors in official work. (8) The comments of analysts last year indicate that much greater difficulty was experienced in securing constant extraction with petroleum ether than with ethyl ether and the experience of the referee and his assistants has been that as extraction proceeds it is necessary to add fresh petroleum ether to secure continuous action of the solvent. (4) Results reported show that the presence of moisture in the sample has a decided influence on the amount of extract obtained from different feeding stuffs and that to some extent at least it affects the determina- tion in cottonseed products. Hence to obtain accurate results drying the sample in hydrogen before extraction seems necessary. (5) Petrolic ether is not of definite or stable composition and results of two analysts on the same sample may be widely at variance owing to the use of petrolic ethers from different sources. This fact is especially important at this time owing to the large demand for gasoline and similar petroleum products. 314 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMIsTs [Vol. I, No. 2 (6) The difference in cost between the two methods on the basis of the cost of solvents is a negligible quantity. (7) The recognition of the petroleum ether method by the association would make official two methods which do not give concordant results and would lead to endless disagreements. If petrolic ether is substituted for ethyl ether as the solvent, comparison with past results will be impossible since the latter has been used almost exclusively for fat determinations in the past. The referee desires to express his appreciation to F. D. Fuller for assist- ance and advice in preparing this report, and to R. E. Nelson, C. Cutler, and J. H. Roop for analytical results. REPORT ON SUGAR AND MOLASSES. By W. E. Cross, Referee. The codperative work on sugar and molasses methods was planned along the lines suggested by previous work. The method involving the use of the Abbé refractometer has already been admitted as a provisional method for molasses, and it was considered desirable to continue the in- vestigations to ascertain whether the methods of determining the water content of sugars with this instrument were reliable, and, furthermore, to determine whether the immersion refractometer could be satisfactorily applied to the determination of moisture in sugar products. The plan also included a comparison of the new direct polarimetric method for molasses with the official Clerget method. Three samples, a raw sugar, a centrifugal molasses, and a blackstrap molasses, were sent to twelve prospective coéperators, but many of these found the pressure of other work too great to allow them to contribute to this work. The following instructions accompanied the samples: INSTRUCTIONS FOR COOPERATIVE WORK. MOISTURE IN SUGAR (1 SAMPLE). (1) Determine moisture in the sugar sample by the official method (Bur. Chem. Bul. 107, Rev., p. 64). (2) Determine moisture in the sugar sample by following method: Weigh 20 grams of the sugar into a tared flask, and add about 20 grams of water. (An ordi- nary 100 ce. sugar flask with narrow neck is suitable.) Then stopper the flask to prevent evaporation, and dissolve the sugar completely by shaking. The total solids value of the solution is obtained by the Abbé refractometer, and temperature correction applied. 1 Presented by P. F. Trowbridge. 1915] CROSS: SUGAR AND MOLASSES 315 : : 2000 — XY Percentage of moisture in sugar = RHI X = Percentage of total solids of solution. Y = Weight of sugar and water. 3. Determine moisture in sugar by the following method: Weigh out 20 grams of the sugar, dissolve in a 100 ce. flask, and make up the solution to the mark. Read off the refractive value of this solution by means of the immersion refractometer (temperature must be constant and carefully noted). From this value the moisture content of the sugar is obtained by reference to the tables.! It is desirable that the refractometric determinations on sugars should be made with extreme care, very special attention being paid to the temperatures, which must be constant and carefully noted. Small errors in temperature reading will produce serious differences in results. MOISTURE IN MOLASSES (2 samples). (1) Determine moisture in molasses by the official method (Bur. Chem. Bul. 107, Rev., p. 65.) (2) Determine moisture by the immersion refractometer, using the following method. Weigh out 20 grams into a 100 ce. flask, and after dissolving make up to mark. Determine dry substance of this solution by means of the immersion re- fractometer. If the tables are not available, give refractometer reading and tem- perature. POLARIZATION OF MOLASSES (2 samples). Determine the true sucrose content of the samples by the official method (Bur. Chem. Bul. 107, Rev., p. 40). Determine true sucrose content of the samples by the following direct method. Dissolve normal weight of molasses and make up to 100 cc. Transfer 50 cc. of the solution to another 100 ce. flask; add 6.3 cc. of sodium hydroxid solution (36° Baumé) and 7.5 ec. of hydrogen peroxid (30 per cent by weight, 100 per cent by volume). Careful cooling is necessary to prevent a too violent effervescence (ether from a dropping funnel can be used to advantage in preventing excessive foaming). Cool- ing in water or ice is helpful in moderating the somewhat vigorous reaction. After effervescence has almost stopped, immerse the flask in a bath at 55°C. for 20 minutes. Cool the liquid, make slightly acid with acetic acid, and make up to mark. After clarification with dry lead subacetate, filter and polarize the solution. The reading multiplied by 2 gives the percentage of true sucrose in the molasses. Please report all temperatures and concentrations used. It is desirable that the coéperative work should be carried out as soon as possible after receipt of the samples, as deterioration and fermentation are likely to take place if the samples are allowed to stand any length of time. RESULTS OF COOPERATIVE WORK. MOISTURE IN SUGARS. The object of this work was to determine to what extent the refractometric method could be applied to the analysis of sugars. The results are presented in Table 1: 1 International Sugar Journal, 1911, 13: 90; La. Agr. Exper. Sta., Bul. 135, p. 18. 316 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMIsTs [Vol. I, No. 2 TABLE 1. Moisture in sugars. 5 5 LUST MEN EH Reon Shen area See per cent per cent per cent W. E. Cross, New Orleans, La.. 1.280 0.99 1.000 G. H. Hardin, New York, Neon 1.220 (Gy OVW Sse. C. L. Clay, New Orleans, La....... 1.278 0.84 0.960 S. F. Sherwood, Washington, DACe 1] 340 0.46 1.120 C. J. Hough, Washington, D.C.... 1.175 0.46 1.015 AVES EG cspacps hem eae apae Rey eee ieaegs 1.259 0.74 1.024 1 After 14 hours, 1.42 per cent. From the results sent in it would appear that the refractometric meth- ods for sugar are not entirely satisfactory. Itis recommended, however, that this work be tried out again, as there is no theoretical reason why the refractometer should give results which are so different from the dry- ing values. While it may be admitted that very slight errors in read- ing, or in temperature connection, will produce serious errors in the values for moisture in the sugar, and furthermore that the limits of possible error with both instruments are rather wide, it is still thought that, with more experience with the methods, results in better accord with the true values might be obtained. MOISTURE IN MOLASSES. Unfortunately, very little coéperation was offered in this part of the work. The results obtained, however, as will be seen from Table 2, were promising, and made it desirable to continue the work in a succeeding year. This method is distinctly the quickest and most convenient for a dark molasses, so that it is important to discover whether or not accurate re- sults can be obtained in this way. TABLE 2. Moisture in molasses. FIRST CENTRIFUGAL MOLASSES BLACKSTRAP MOLASSES ANALYST Official Immersion Official Immersion method refractometer method refractometer per cent per cent per cent per cent Wm. E. Cross, New Orleans, La.| 24.800 25.00 22.00 21.30 C. J. Hough, Washington, D.C.) 27.035 27.50 26.10 24.75 S. F. Sherwood, Washington, Di Co ee Rae eee DEAE Ao Oe 125 .890 27.65 224.24 25.35 ASVEL AG Or. crohns aon 25.900 26.71 24.11 23.80 127.10 in 14 hours. 226.22 in 14 hours. 19165] RATHER: TESTING CHEMICAL REAGENTS 317 POLARIZATION OF MOLASSES. The results obtained with the determination of sucrose by direct polari- zation were gratifying, and show that the method is well deserving of urther study, especially as these results only confirm the satisfactory re- sults obtained in the referee work of last year. The method is much quicker and simpler to operate than the Clerget determination, so that if it were proved reliable it would be of much service to the sugar chemist. The results of the codperative work are presented in Table 3. TABLE 3. Polarization of molasses. FIRST CENTRIFUGAL MOLASSES BLACKSTRAP MOLASSES ANALYST Rm eT Son lee eS ae Official method | Direct method | Official method | Direct method per cent | solution|per cent | solution| per cent | solution|per cent | solution Wm. E. Cross, New Orleans, d TLE, A Cae Ra Bn a 43.03] 3} AZO |p eat 29N0 NI 27040l as H. Z. E. Perkins, New Or- ROSS login. sa erecta 43-36) ce 43,05) -. 29.2 Ss 30.05 G. H. Hardin, New York, INI OR eel cc a Settee ek 43.75) 3 43.60} 4 29 .54 es 30.80 S. F. Sherwood, Washing- iG), 1D sil CS Re ruin een 43.57} .. 47).12)| "=. 23 .64 x 21 dD C. ai Hough, Washington, BCL. a 43 512)) 22 43.05} .. 28 .94 Be 28 .40 C. Fis Clay, New ‘Orleans, Da orarat fol cransysie hase Seevataye shalt; 41.00) .. 42.44) .. |141.0 52 ih2 044 PAVETR EC . cccs cece seen] 42690 42.69 29.064; .. 28 .84 1 Omitted from average. RECOMMENDATIONS. It is recommended that, (1) the refractometric methods for determin- ing the moisture content of sugars, (2) the determination of the moisture content of molasses by means of the immersion refractometer, and (3) the direct determination of sucrose in molasses, be further studied. REPORT ON TESTING CHEMICAL REAGENTS. By J. B. Ratusr, Referee. The work conducted by the referee differs somewhat from that taken up by the Committee on Testing of Chemical Reagents. The committee seems to have studied the impurities in the various reagents without re- gard to the probable effects of impurities on any particular determination. The referee has limited his work to a few reagents and has tested only for these impurities likely to be present which may affect the results of spe- cific determinations, or whose effect on the results is not known. 318 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMIsTs [Vol. I, No. 2 The purity and strength of crude caustic soda, molybdic acid, commer- cial citric acid, and ethyl ether have been studied in regard to their suita- bility for the determination of nitrogen, phosphorie acid, insoluble phos- phorie acid and ether extract respectively. Since it could not be foretold what methods would be needed, it was considered best to conduct all the analytical work for this year in this laboratory, with the view of testing codperatively next year, methods for the determination of those impurities whose nature and amount may justify it. About ten samples each of crude caustic soda, molybdic acid, and com- mercial citric acid, were sent in by the following codperators, and were examined with the results given in this report: R. C. Thompson, Fayette- ville, Ark., Wm. G. Gaessler, Ames, Ia., A. E. Vinson, Tucson, Ariz., W. J. Jones, jr., Lafayette, Ind., Wm. P. Headden, Ft. Collins, Col., T. O. Smith, Durham, N. H., W. A. Withers, West Raleigh, N. C., J. T. Wil- liard, Manhattan, Kans., Wm. Frear, State College, Pa., E. Van Alstine, Urbana, Ill., H. D. Haskins, Amherst, Mass. CRUDE CAUSTIC SODA. Methods for the determination of sodium hydroxid and sodium carbonate in crude caustic soda. While the presence of a small amount of sodium carbonate is probably not objectionable in nitrogen determinations, a large amount tends to cause frothing, and for that reason it was thought desirable to find a satis- factory method for its determination. The amount of sodium hydroxid in crude caustic soda should be known because samples are likely to vary considerably in their content of this substance, and, therefore, a satur- ated solution, or one with an arbitrary percentage of crude caustic soda, is likely to vary greatly in its content of sodium hydroxid. (1) Quantitative determination and determination of sodium carbonate content (Krauch-Merck).—Dissolve 2 grams of crude caustic soda in 200 cc. of water and titrate 20 ec. with fifth-normal hydrochloric acid in the cold, using phenolphthalein as an indicator. When the red color has disappeared, read the burette, add a drop of Methyl Orange, and titrate further until the color changes to red. Subtract the last reading from the first and multiply by 4. The result is percentage of so- dium hydroxid. Multiply the last reading by 10.6; the result is percentage of sodium carbonate. (2) (Sutton).—Place 100 ce. of the above solution in a 200 ec. flask, add 10 ee. of 10 per cent barium chlorid, heat to boiling, shake well, and let cool. Make up to volume, filter through a dry filter into a dry flask and titrate 40 ec. with fifth-normal hydrochloric acid and phenolphthalein. Multiply the reading by 4 to get percentage of sodium hydroxid. Subtract the reading from the sum of the two readings ob- tained in (1) and multiply the remainder by 5.8. The result is the percentage of sodium carbonate. 1915] RATHER: TESTING CHEMICAL REAGENTS 319 (3) Determination of sodium carbonate by precipitation as barium carbonate. (Sutton).—Dissolve 2 grams of crude caustic soda in water, add barium chlorid so- lution and heat to boiling. F{lter and wash with hot water; dissolve the barium carbonate from the filter with 15 ce. of fifth-normal hydrochloric acid, wash well and titrate the filtrate, after boiling to expel carbon dioxid, with tenth-normal sodium hydroxid and phenolphthalein. Multiply fifth-normal acid consumed by 0.53; the product is percentage of sodium carbonate. (4) Determination of sodium carbonate by direct titration (Referee’s modification). —Dissolve 2 grams of crude caustic soda in water and titrate with approximately twice-normal hydrochloric acid and phenolphthalein until the color fades. Titrate another 2-gram portion, using 0.5 cc. less of the twice-normalacid. Now titrate this solution with fifth-normal hydrochloric acid until the color fades; read the burette, add 2 or 3 drops of Methyl Orange, and titrate until the color changes. The number of cubic centimeters of fifth-normal acid required to change the color of the Methyl Orange multiplied by 1.06 gives the percentage of sodium carbonate. TABLE 1. Determination of sodium hydroxid and sodium carbonate in crude caustic soda by different methods. SODIUM EYDROXID SODIUM CARBONATE 5 Gs} LABORA- 2 5 8 2 & TORY DESCRIPTION S ao S a No. sa | 8, [e688 | Sa | 8, | Sa | Se 3 8 | a53| 3 8 3 5 g g |) Gen q 5 5 3 4 a | PSS4|| B a 4 per cent |per cent |per cent |per cent |per cent |per cent |per cent 7022 Greenbanks (98 per cent)............]| 93.2 91.6 ual 2.4 305 BAG 7023 do 93.4 91.8 aut 2.1 6.0 2.0 7024 do 93.3 92.0 3.2 2.8 5.0 1.6 ROZO Me MEerculesinyeaedtseltiercciaete aces etnciae 77.0 75.6 ee 3.2 2.5 5.5 2.8 7036 | Greenbanks (98 per cent)............| 93.0 94.4 92.2 2.7 3.5 4.5 2.3 7043 | Baker and Adamson’s Electrolytic (Q8ipericent).ccF.20sic00 Jeisis ciereisie'e 91.6 92.8 ation 3.2 1.6 4.4 1.3 7134 Greenbanks (98 per cent)............ 93.2 92.8 91.6 Se) 3.2 4.6 Pyal 7137 | Electrolytic (98 per cent)............ 91.6 91.6 91.2 3.2 3.2 4.8 2.4 7140 | Henry Heil Chemical Co. Electro- lytic (98 per cent)................. 92.4 92.4 91.8 3.2 3.2 4.4 7143 Greenbanks (98 per cent)............ 94.8 95.2 95.0 3.2 227 3.8 All of the samples except one contained more than 90 per cent of sodium hydroxid and that one (No. 7025) contained only about 75 per cent. The determinations of sodium hydroxid by Methods (1) and (2) leave much to be desired in the way of agreement. In both of these methods 0.1 cc. is equal to 0.4 per cent of sodium hydroxid, and it is evident that the error in titration could easily be more than 0.25 ce., equivalent to 1 per cent. While great accuracy is not necessary in this work, the limit of error could be cut in half by doubling the amount of sodium hydroxid solution taken for titration. This has been done in a few cases (see Column 5 in table). The results, however, do not differ materially from those by Methods (1) and (2). 320 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMIsTs [Vol. I, No. 2 Method (1) modified so as to double the amount of solution taken for titration, appears to be more convenient and equally as reliable as Method (2) and should be studied further. Methods (1) and (2) for the determination of sodium carbonate gave discordant results, both among duplicate determinations and determina- tions by the different methods. This is due largely to the fact that 0.1 ce. of fifth-normal acid is equal to 0.53 per cent sodium carbonate; the error of titration could easily affect the result 1 per cent or more, and these methods are, therefore, considered too rough for even approximations. Results by Method (8) are much higher than by (1) and (2). With Method (4) the samples of sodium hydroxid were by this time nearly exhausted and had no doubt taken up water from the air. The results may not be strictly comparable with those obtained by the other methods. The following results on No. 7023 were obtained with this method: 1.8, 1.9, 2.0 per cent. This method should be studied further. The amount of sodium carbonate found in these samples, does not ap- pear to be excessive, but further work is necessary before a limit to the amount allowed could be set. Conditions affecting the determination of nitrogen in caustic soda. (A) Effect of the amount of sodium hydroxid on the apparent nitrogen.— This work was undertaken to see if it would be desirable to determine the percentage of nitrogen directly, a large amount of the caustic soda being used. The methods used were as follows: (1) Put 100 grams of caustic soda in a Kjeldahl flask with a little granulated zine and distill into a receiver containing 5 ee. of fifth-normal hydrochloric acid. Titrate the excess of acid with tenth-normal ammonia and cochineal and report cubie centimeters of acid consumed. (2) Proceed as in (1) but use 40 grams of caustic soda (this weight is contained in the volume of solution generally used in nitrogen determinations). TABLE 2. Effect of the amount of sodium hydroxid on apparent nitrogen. FIFTH-NORMAL HYDROCHLORIC ACID CONSUMED LABORATORY NO. Method 1 Method 2 ce. cc. 1.80 0.15 0.65 0.10 0.50 0.20 0.40 0.05 0.45 0.30 0.55 0.10 1.80 0.15 1.05 0.20 0.60 0.20 0.80 0.05 IAVELABC Rass dae h mace eee 0.86 0.15 1915] RATHER: TESTING CHEMICAL REAGENTS 321 The acid consumed when 100 grams of caustic soda were taken, was in most cases much higher than when 40 grams were taken, averaging 0.86 cc. for the former and 0.15 ce. for the latter. This would correspond to an error in determination, when 0.7 gram is taken for analysis, of 0.344 and 0.060 per cent of nitrogen respectively. If the 0.15 cc. represented ammonia, in the cases where 40 grams of caustic soda were used, then the acid consumed by the distillate from 100 grams should be 23 times 0.15 or 0.38 ce. The result actually obtained, however, was 0.86, more than twice as much. It appears, therefore, that some of the caustic soda was carried over mechanically, thus causing an increase in the apparent ni- trogen. The useof 100 gramsof caustic soda in the determinates of nitro- gen is undesirable and leads to erroneous results. (B) Effect of the addition of sulphuric acid on the apparent nitrogen.— Since the use of 100 grams of caustic soda gave erroneous results, probably due to spitting of the alkali solution, it appeared possible that this same effect might be found when only 40 grams were used. In order to test this point and more nearly duplicate the conditions of an actual determination of nitrogen the following methods were used: (1) Proceed as in Method (2) given under (A). (2) Proceed as in (1), but add a solution containing 23 cc. of concentrated sul- phuric acid to the soda before distillation, making sure that the solution is alkaline by means of phenolphthalein after the addition of the acid. TABLE 3. Effect of addition of acid (concentrated sulphuric) on apparent nitrogen. 40 GRAMS OF SODIUM HYDROXID LABORATORY NO. Method 2 apn 23 ce. ot concentrated sulphuric cc. fifth-normal acid used cc. fifth normal acid used CUPS es caeaslens CURES Ce ees rca 0.13 0.15 cea er Ser terrenvsinir ose saleraliayens rai 0.20 0.10 (ees oo Peach eeiicin c OO eee 0.15 0.20 TAOS 3 te GRE oe D cee mee 0.13 0.05 OSD MPR Pies sae s eieanon eect ae 0.15 0.30 OA SMe ceesE final fo. atsicg ages eteps¥3, sieeve: wang 0.13 0.10 Eset acy ote oes ois Pw te see ategeseiore 0.20 0.15 CUS Vie Bocce aoe ae OTe Cee 0.17 0.20 TTD) ei Se aire a ae a 0.17 0.20 BLAS Re Petcrishattesn hesjc. toeunys peers eee 0.20 0.05 FAVELA LCs Tectia: ss ee isaac cee ticles 0.16 0.15 The amount of acid consumed by the distillates in thetwo methods showed no wide variations and averaged practically the same, 0.16 cc. and 0.15 ec., for Methods (1) and (2) respectively. This would corre- 322 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMIsTs [Vol. I, No. 2 spond to an error in determination, when 0.7 gram of substance is taken, of 0.064 and 0.060 per cent of nitrogen. From these results it appears that the use of sulphuric acid has no effect on the results when 40 grams of caustic soda are used; its use is therefore considered unnecessary. (C) Effect of redistillation of distillate on the apparent nitrogen.—To test more conclusively whether the apparent nitrogen in the caustic soda was due to such causes as spitting, errors of solution and mensuration, six portions of 75 ec. each of a stock solution of caustic soda (Greenbanks), containing about 40 grams of sodium hydroxid, were distilled with a little zine into fifth-normal hydrochloric acid after dilution with water, and the amount of acid consumed determined. (1) Six portions of the same solution were distilled into distilled water, the distillate returned to a clean Kjeldahl flask and again distilled, without the addi- tion of anything, into fifth-normal hydrochloric acid. (2) The same amount of fifth-normal hydrochloric acid used in all the receivers (5 ec.) was measured out and diluted to about the same volume (400 ce.) as the usual distillate. The usual volume of cochineal solution was then added and the solution titrated against tenth-normal ammonia. The amount of ammonia required to neutralize the acid was divided by 2 and subtracted from the volume of acid taken. This correction for errors of solution, mensuration, and personal factor (depth of tint of color to which titrated, etc.) was applied to the average results. TABLE 4. ‘Effect of redistillation of distillate on apparent nitrogen. METHOD 1 METHOD 2 LABORATORY NO: (Distilled once) (Redistilled distillate) ce. fifth-ncrmal acid used ec. fifth-normal acid used LEENA Gao ON Ctimpei caer ee anIce 0.30 0.40 Director apetars foveal obser ave ae ene aie Sareea 0.45 0.40 BS Oe EEOC Oe GO RE aera 0.55 0.30 CTO ace erie Bete e isiceocd eee 0.45 . Rota AORN Pearce c ropwiar tatters et aie rece are ene 0.55 siete (Rees fern Oe a Caer Gea 0.45 0.50 /ARTGIENRO: oo sane adoogoob090000 2 0.46 0.40 Corrected for errors of solution, etc. (0.20 ce.), Method (1), 0.26; Method (2), 0.20; error in determination when 0.7 gram is taken. Method (1), 0.104; Method (2), 0.080 per cent of nitrogen. The average amount of acid consumed by the solution which was dis- tilled once (1) was 0.26 cc. (corrected), while that consumed by the redis- tilled product (2) was 0.20 ce. This would correspond to an error of 0.104 and 0.080 per cent nitrogen respectively when 0.7 gram is taken for analy- sis. The differences are small and well within the limit of error. The point, however, deserves further study. It will be noted that the cor- 1915] RATHER: TESTING CHEMICAL REAGENTS 323 rection applied was —0.20 cc., that is, the tenth-normal ammonium hy- droxid required to neutralize 5 cc. of fifth-normal hydrochloric acid was only 9.80 ce. instead of 10 ce. as required by theory. This large difference was not found in our previous work. The average of a number of tests by two analysts gave the figure, —0.03. MOLYBDIC ACID. Methods for the determination of molybdic anhydrid. The lead molybdate method recommended by Krauch-Merck was used first, but this proved very tedious and a slight modification made by the referee was used. Attempts were made to devise a suitable volumetric method for the determination of molybdic anhydrid. It was found that comparable results could not be obtained in a volumetric determination in the presence of the necessary excess of phosphoric acid, unless the con- ditions as to acidity were exactly the same. Ammonium nitrate solution was accordingly used instead of nitric acid and ammonium hydroxid. The methods as finally used are as follows: (1) (Krauch-Merck).—Dissolve 0.5 gram of molybdic acid in 50 ce. of water and 1 ce. of ammonium hydroxid, heating gently. Acidulate with 5 ec. of acetic acid, dilute to 200 ce. with water, heat to boiling and add a filtered solution of 1.5 gram of crystallized lead acetate in 20 cc. of water. Boil several minutes, stirring con- stantly. Collect the precipitate on a filter dried at 100°C. and wash with boil- ing water. Dry the precipitate to constant weight at 100°C. and ignite a portion. PbMoO, X 0.39247 = Percentage of MoQ3;. (2) (Referee’s modification of (1).—Dissolve 0.5 gram of molybdiec acid in 50 ec. of water and 1 ec. of ammonium hydroxid, heating gently; filter if necessary; acidu- late with 5 ce. of acetic acid, dilute to 200 ec. with water, heat to boiling, and add a filtered solution of 1.5 gram of crystallized lead acetate in 20 cc. of water. Boil several minutes, stirring constantly. Allow to settle a minute or two and decant through a Gooch crucible with a fairly thick felt, which has previously been ignited and weighed. Wash by decantation 10 times with 50 cc. portions of boiling water, allowing about a minute for the precipitate to settle each time. Transfer the precipi- tate to the Gooch and remove the water by suction. Ignite the precipitate with a blast lamp without further drying, cool, and weigh. The ignited precipitate is Pb MoO,. Pb MoO, X 78.494 = Percentage of MoQ3. (3) (Referee).—Dissolve 10 grams of molybdie acid in 15 ee. of ammonium hy- droxid and 27 cc. of water and pour into 55 ce. of nitric acid and 100 ce. of water in a 200 ce. flask slowly and with constant shaking. Cool and make up to volume with water. Allow to settle overnight and filter. Add a solution containing 0.040 gram of P20; (sodium phosphate) to 75 cc. of water in a beaker, and add a solution containing 7.5 grams of ammonium nitrate. Heat to 65°C. and add 20 cc. of the molybdate solution. Complete as for volumetric phosphoric acid. The number of cubic centimeters of standard potassium hydroxid consumed multiplied by 2.6941 gives the percentage of MoO; in the molybdic acid. 324 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 TABLE 5 Molybdic anhydrid in molybdic acid by various methods. LABORA: Aeon arta il | eon allesees percent | percent | percent 7019 | In 50 pound kegs, purity unknown......... 85.8 oe \ waite 7O201 EAI bauer te oe eee eee 99.0 1 108 } 96.25 7021 || Himer-and-Amend.(Saipen Cont)sqe sss... 91:7 jl. eae } 94.36 zoos | Merck, Molybele anhyatide (99.5-100er |¥499.5 |{ 190-08 | ton.2 7038 | Mallinkrodt, Molybdic acid............... 10024 ee } 97.65 7039 | Molybdic acid (10 years old),.....-.-:..-.| ----- oe } 87.90 7042 | Merck, Molybdic anhydride ..............| ..... cope \ 94.29 7135 | Marquard, Molybdie acide. p..........----| «+--+. i516 \ 78.07 7138 | Molybdic acid (85 per cent).............--| ----- eae } 88.30 rT MIP eres WN Gen a RRevatL coca Wecaecoucsearl|| secs a0 } 02.89 7144 | B. & A. Molybdic acid, ec. p. (99.9 per cent).| ..... see } 97.33 The results by the Krauch-Merck method (1) do not differ greatly from those by the method as modified by the referee; (2) the only difference in the methods is in manipulation. The original method requires practi- cally 2 days to complete a determination, while a single determination by the modified method can be completed with good suction in 2 hours, and six determinations have been completed in 3 hours. The results by the volumetric method, (3), vary from 4.8 per cent below to 1.2 per cent above the average results of Method (2). In no case are they the same. The factor 2.6941 used in calculating these results was obtained by dividing the average of the results by Method (2) by the average of three closely- checking determinations of standard potassium hydroxid consumed by the phosphomolybdate precipitate, and using the average of the figures thus obtained. It is evident that, if there is any constant relationship between the potassium hydroxid consumed by the molybdate precipitate under the conditions, and the content of molybdic anhydrid in the molybdic acid, results obtained by the use of the factor would at least approximate those by Method (2). This is not the case; the results are invariably low when the samples are high in molybdie anhydrid and high when they are low in it. The volumetric method (3) can not, therefore, be used as a reliable method for the determination of molybdic anhydrid in molybdie acid. Method 1915] RATHER: TESTING CHEMICAL REAGENTS 325 (2), the modified Krauch-Merck method, is more rapid than either of the other two, gives closely-checking results and seems to be the most desir- able of the three. It will be noted that by Method (2) some of the samples of molybdic acid contain over 100 per cent of molybdie anhydrid. These samples were described as molybdie anhydrid (MoO), and all had a bluish tint due to the presence of other oxids of molybdenum (Krauch-Merck, p. 18). These oxids, MoO, and MosQ3, contain more molybdenum than MoQs, and this probably accounts for the higher results. Molybdie acid, H:Mo0Ox,, contains about 88.9 per cent MoO;, and HeMoO,H.O contains about 80.0 per cent. The ordinary 85 per cent molybdie acid is probably a mixture of these two compounds, but the content of sample No. 7135, which is described as molybdie acid e. p. is 75.16 per cent of MoO;. This figure is too low to be accounted for by this hypothesis. The samples examined varied from 75.16 per cent to over 100 per cent of molybdic anhydrid. Sample No. 7021, claimed to be 85 per cent molybdic anhydrid, contained 92.2 per cent, while Sample 7042, claimed to be molybdie anhydrid, contained only 93.1 per cent of molybdic anhy- drid. Molybdate solutions prepared by the official method from 7144 would have a precipitating power 33 per cent greater than 7135. In the official methods H.MoO, (85 per cent) is probably meant by “molybdie acid,’’ but nearly half of the codperators seem to be using the anhydrid, Mo0O;. PHOSPHORIC ACID IN MOLYBDIC ACID. Total phosphoric acid, phosphoric acid in the official molybdate solu- tion, and sulphates were determined. The methods follow: (1) Total phosphoric acid.—Dissolve 1 gram of molybdic acid in a few cubic cen- timeters of ammonium hydroxid, dilute to about 75 cc. and acidify with nitric acid. Heat at 65°C. for 15 minutes and complete as for volumetric phosphoric acid. (2) Error in phosphoric acid determinations caused by phosphoric acid in molybdate. Prepare the solution as described in (3) under Methods for the Determination of Molybdic Anhydrid. Dilute 50 ec. with about 25 ce. of water, digest and complete as for volumetric phosphoric acid. Calculate to phosphoric acid on the basis of 0.2 gram taken for analysis. (8) Sulphates in molybdic acid.—Dissolve 1 gram of molybdie acid in 2 ec. of ammonium hydroxid, heating gently, dilute to 150 ce. and acidify with hydrochloric acid, and heat to boiling. Add barium chlorid solution and allow the precipitate tosettle. Filter and complete as usual. 326 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMIsTs [Vol. I, No. 2 TABLE 6. Phosphoric acid and sulphates in molybdic acid. METHOD (2)! ERROR IN a A METHOD (1), TOTAL OREO ROR METHOD (3) R Al 4 PHOSPHORIC ACID DETERMINATIONS SULPHATES LABORATORY NO. per cent per cent OLOM Peter ets ee aes 0.00 None None US: SRR OO Aero ntae 0.00 None None TROPA Serer er ee eee etn oe Gs 0.00 None None OLS tee tiontcnis aah ae oro 0.00 None None MOSS iret ttyl este eee 0.00 None None TUS Weaeieaanlat ome suerte bists 0.00 0.08 None kW Oni Se Brea rc SB oo5.s 0.00 0.04 None (S54 fees ee aeepemeertee 0.00 0.04 None WSS ssc Orn cee C ER Eee 0.01 0.01 None ML AI 3 3/5 5A 2 me octreies Shes oe 0.00 0.04 None TAA SCOR ER aire Soo see eis 0.01 0.05 None 1 No visible precipitate formed in any case, those marked ‘‘ None”’ were not completed. Only two samples of molybdic acid contained an appreciable amount of phosphoric acid and these only 0.01 per cent. In the determinations of phosphoric acid in the official molybdate solution no visible precipitation formed in any case and a number of the determinations were not carried further. The figures given under Method (2) represent merely errors of titration. The presence of sulphates was considered because of their probable effect on the determination of molybdie anhydrid by the lead method. No visible precipitates formed, and the determinations were not completed. It appears that the errors introduced by phosphoric acid in molybdie acid are insignificant. The question of the effect of sulphates on the determination of molybdie anhydrid in molybdie acid should be studied further. CITRIC ACID. The samples of citric acid were tested for ash and oxalic, tartaric, and sulphuric acids, and sugars. The following methods were used: Methods. Total sugars in citric acid.—Dissolve 10 grams of citric acid in 25 ec. of water and add 5 cc. of concentrated hydrochloric acid. Heat in a water bath to 69°C. in about 3 minutes and maintain at this temperature about 7 minutes, making a total heating of 10 minutes. Remove from bath, cool rapidly to room temperature, neu- tralize with sodium carbonate, determine reducing sugars by Allihn’s method, and determine the reduced cuprous oxid by Low’s volumetric method. Ozxalic and tartaric acids —Dissolve 10 grams of citric acid in 20 ce. of water and add 5 ce. of 1 to 2 potassium acetate solution and 50 ce. of 85 per cent alcohol. No turbidity should be produced, nor should a crystalline form appear within 2 hours. Tartaric acid and sugar (qualitative) —Grind 1 gram of citric acid and 10 ce. of sulphuric acid together in a porcelain mortar previously rinsed with sulphuric acid. 1915] RATHER: TESTING CHEMICAL REAGENTS 327 When this mixture is heated in a test tube for an hour in a boiling water bath, it acquires at most a slight yellow color, but no brown color should develop. Ash.—Ignite 10 grams in a tared porcelain crucible and weigh. Sulphuric acid.—Dissolve 10 grams in water and acidify with hydrochloric acid. Add barium chlorid and complete in the usual manner. Oxalic acid.—Neutralize a solution containing 10 grams of citric acid with am- monium hydroxid and acidify strongly with acetic acid. Add calcium chlorid and let stand a few hours, filter and wash. Ignite the precipitate and weigh. Calculate the amount of the oxalic acid present from the weight of the lime precipitate. It should be remembered that calcium citrate precipitates from the above solution when it is only weakly acid, when heated, or when it stands overnight. Tartaric acid (qualitative) —Dissolve 0.5 gram of the sample in 10 ce. of water and add, drop by drop, to 15 cc. of lime water. No turbidity should be produced. In the samples examined, the ash contents varied from 0 to 0.05 per cent and averaged 0.02 per cent. The amount of ash constituents intro- duced into 100 ce. of citrate solution would vary from 0 to 0.0093 grams. The weight 0.0093 is negligible when the relative amounts of soluble bases brought into solution by the citrate are considered, and for that reason no further examination of the ash was made. No oxalic acid was found in the samples examined. The amount of total sugar, calculated as dextrose, varied from 0 to 0.029 per cent and averaged 0.005 per cent. All but two of the results for sugar are, how- ever, within the limit of error. The maximum amount of sugar introduced into 100 ce. of citrate solution by the citric acid examined would be 0.0056 gram. While the effect of sugar is not known to the referee, the above amount is considered too small to justify an investigation. In only one sample, No. 7139, was there any appreciable amount of sul- phuric acid. The percentage was 0.014 and would correspond to 0.0026 gram in 100 ce. of citrate solution. Much more of the sulphates than this would quite likely be dissolved from the fertilizer sample by the ammonium citrate solution. Qualitative tests for oxalic and tartaric acids, and for tartaric acid, gave negative results in all cases. In the test for tartaric acid and sugar a few samples gave a faint test, but this could be due to small fragments of insoluble organic matter like excelsior or paper, which are possible accidental contaminants. In view of the results presented above, it appears that the purity of the samples of citric acid examined is as high as could be expected, and quite sufficient for fertilizer control work. The subject should be studied further. ETHYL ETHER. Four samples of ethyl ether, distilled over sodium (Kalbaum’s) were examined for matter nonvolatile at 100°C. The method was as follows: 328 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMIsTs [Vol. I, No. 2 Method. Evaporate 100 ce. in a tared platinum dish to dryness on a steam bath. Dry at 100°C. to constant weight. The acidity and ash were determined in the one sample which was found to be badly contaminated. The results are shown in Table 8. TABLE 7. Examination of commercial citric acid. roman | oo. QUALITATIVE TEST J - 1G bs mrocce SOURCE OF SAMPLE ASH Saad Sere See Oxalie |Tartaric ? NO. = DEX- (SO) and | acid |Tartaric TROSE tartaric] and acid acid sugar per cent per cent | per cent COLD ||| Maltinkrod teria pieiisciecie raisin sieicier = 0.04 | none 0.000 | trace! | none | trace | none 7016 Bought in kegs, unknown purity....}| 0.00 | none 0.000 | none none trace none 7017 | Eimer and Amend..................- 0.01 none 0.005 | none | none trace none 7018 do 0.01 | none 0.000 | none none | trace | none 7037 | Eimer and Amend (U.S. P.).......- 0.05 | none 0.029 | trace! | none | trace | none 7040 Ohvearsiold see ssc yereeine tees 0.01 | none 0.003 | tracel | none | none | none 7041 | Baker and Adamson.................| 0.02 | none 0.003 | tracet | none | none | none 7136 Pfitzer (99 per cent)................- 0.02 none 0.003 | trace! | none none none 7139 | Eimer and Amend...................] 0.03 | none 0.001 | 0.014 | none | none | none M42 | Sargent and! Oon.ceees ecient saeieices 0.01 none 0.001 | trace? | none none none 1 About 0.001 per cent. TABLE 8. Amount of various impurities in ethyl ether distilled over sodium. = ACIDITY AS SAMPLE NO. SOURCE OF SAMPLE SO Onea ASH Sue RCE aero gram in 100 ce. | gram in 100 cc. | gram in 100 cc. gram tn 190 cc. 1 Kalbaum’s..... 0.0600 0.0070 0.37 0.60 to 1.50 2 do OMONG. WW doecce cere 0.02 to 0.04 3 do Q20026io A ener Mice 0.03 to 0.07 4 do OL002250 st Seen: rece 0.02 to 0.06 The matter nonvolatile at 100°C. varied from 0.0015 gram to 0.0600 gram per 100 cc. One sample contained 0.0070 gram of ash in 100 ce. and an acidity corresponding to 0.37 per cent of sulphuric acid. The ether-blackened feed samples on which fat determinations were made and the ether residue gave off whitish fumes on ignition. In this laboratory from 20 cc. to 50 ce. of ether are taken for a determination of fat and the error introduced by the matter nonvolatile at 100°C. would vary from 0.02 to 0.07 per cent to 0.60 to 1.50 per cent. The lower figures are of slight significance, but the higher ones deserve serious attention. The method for the determination of ether impurities nonvolatile at 100°C. should be studied further, together with the question of the presence of alcohol and water, and possibly other impurities. 1915| BACON: TANNIN 329 RECOMMENDATIONS. It is reeommended— (1) That the lead molybdate method recommended by Krauch-Merck for the determination of molybdie anhydrid in molybdic acid, as modified by the referee, be studied coéperatively, together with the methods for the determination of nitrogen in sodium hydroxid and impurities of ether nonvolatile at 100°C. (2) That the effect of the presence of sulphates on the accuracy of the lead molybdate method for the determination of molybdie anhydrid be studied, and that the nature and amount of the impurities in citrie acid and ether, methods for the determination of sodium hydroxid and sodium carbonate in crude caustic soda, and the amount of phosphoric acid in molybdie acid, be studied further. REPORT ON TANNIN. By C. B. Bacon, Referee. Since there has been no collaborative work on tannin analysis this past year, and, therefore, no report, it might be well worth while to call the attention of the members of the association to the more recent changes in the methods of analysis of the American Leather Chemists Association, which is composed of approximately 125 active members, who are vitally interested in the matter of tannin analysis. The points wherein their methods differ from those of the Association of Official Agricultural Chem- ists are briefly as follows: 1. Extraction.—In extraction the tendency is toward narrowing the per- missible limits of tannin content from 0.35 to 0.45, as it now is in the As- sociation of Official Agricultural Chemists method, to 0.375 to 0.425 grams per 100 ee. 2. Cooling —The American Leather Chemists Association permits the rapid cooling of the solution to analytical temperature of 20°C. 8. Filtration—The temperature of filtration is kept at 20°C. 4. Evaporation and drying.—The use of the combined evaporator and dryer is specified and the time recommended as 16 hours. Aside from these points the methods are practically identical. Inasmuch as extracts which contain wood pulp liquor are presented for analysis according to the official methods, it would seem advisable that the better known qualitative methods for the detection of wood pulp liquor be included so as to be available for use. Probably the best known is the modification of the lignin test by Procter and Hirst,! which is as follows: 1 J. S.C. 1. 1909, 28: 293. 330 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMIsTs [Vol. 1, No. 2 To 5 ce. of the extract solution, which should be of about the ordinary strength employed for analysis, 0.5 ec. of fresh, pure aniline is added, and the whole is well shaken, and 2 ec. of concentrated hydrochloric acid is then added to the mixture. With all ordinary extracts this has the effect of immediately clearing the turbidity caused by the aniline, and a perfectly transparent solution results, but where pine wood extract is present, even in comparatively small quantity, a precipitate is produced which gradually rises to the top of the liquid. Heating is not neces- sary, and on the whole not desirable, though it sometimes increases the rapidity of the separation of the precipitate. The reaction, however, is immediate, and any slight turbidity which arises after considerable standing should be disregarded, as it sometimes occurs in the case of unmixed extracts, probably from minute traces of ligneous matter. A more recent method by Hayes! is as follows: A mixture is made of 40 ce. of a 2 per cent gelatin solution and 30 ce. of glacial acetic acid. Seven ce. of the reagent is added to 10 cc. of the solution to be tested, this latter being of regular analytical strength. If wood pulp liquor is present, a precipitate is formed which persists even on heating. Should the presence of wood pulp liquor be shown by these tests, an approximately quantitative determination may be made, if it is in consid- erable quantity, by using a modification of the Loewenthal method, as described by Procter.? A modification of a method by Hinrichsen for the determination of tannin and gallic acid in inks has been found useful in the Bureau of Chemistry. The procedure is as follows: Place 10 grams of ink in a 100 cc. separatory funnel, add 10 ce. of 20 to 25 per cent hydrochloric acid; shake five times with 25 cc. portions of acetic ether. Unite the acetic ether shakings in a 200 to 250 ec. separatory funnel and shake with 10 ce. portions of half saturated potassium chlorid solution until the aqueous layer does not give a reaction for iron. Evaporate the acetic ether solution of the tannin and gallic acid to dryness, preferably in vacuo, or at the lowest possible temperature. Wash the residue by means of a small amount of water into a weighed evaporating dish and evaporate to dryness on the steam bath. Dry the residue at 105°C. for 2 hours and weigh as tannin and gallic acids. It is suggested that in the near future the methods of the Association of Official Agricultural Chemists be tested to show whether they are su- perior to those of the American Leather Chemists Association in order to get an even closer agreement between the two methods, so that all may profit by the collaborative work that is being done by that association. The secretary announced that invitations had been received from a num- ber of associations and organizations in St. Louis, for this association to meet in St. Louis in 1914 and from San Francisco for this association to meet in San Francisco in 1915. The secretary was requested and author- ile labs (Gn “lS IBY BS 7h 2J.S8. C. I. 1909, 28: 294. 1915] TROWBRIDGE: REPORT OF COMMITTEE B 331 ized to decline the former, expressing the appreciation of the association, and to reply to the Jatter that it would receive careful consideration at the next meeting. L. F. Kebler announced a meeting of the local branch of the American Pharmaceutical Association that evening and invited all members and visitors at the convention to attend. REPORT OF COMMITTEE B ON RECOMMENDATIONS OF REFEREES. By P. F. Trowsrinae, Acting Chairman. (Dairy products, foods and feeding stuffs, sugar, water in foods, organic. and inorganic phosphorus in foods, separation of nitrogenous bodies, test- ing chemical reagents, tannin, medicinal plants and drugs.) DAIRY PRODUCTS. It is recommended— (1) That in view of the distinct advantages of the copper sulphate method of preparing milk serum, chief of which are the rapidity with which the serum can be obtained and the narrower range of readings given by whole milk, the referee for the ensuing year consider this method for the purpose of having it adopted as an optional provisional method in 1914. Approved for final action in 1914. (2) That the referee for the ensuing year study the Harding-Parkin method for fat determination (J. Ind. Eng. Chem., 1913, 5: 131) in com- parison with the present official and provisional methods. Approved. (3) That the subjects now under consideration be given further study. Approved. (4) That in Bulletin 107, Revised, page 122, after the paragraph on Cream, the following paragraph be inserted under the heading “Con- densed Milk (unsweetened) :” Dilute 40 grams of the homogeneous ma- terial with 60 grams of distilled water, proceed as directed under ‘ Milk,”’ and correct the results for dilution; and the word ‘‘Sweetened’’ be in- serted before the word ‘‘Condensed” in the subsequent heading. Referred to the incoming referee for a report next year. FEEDS AND FEEDING STUFFS. It is reeommended— (1) That the recommendation of the referee in 1911 (Cir. 90, p. 8) that the association recognize the petroleum ether method for determining fat in cottonseed products, be not adopted. Approved. 332 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMIstTs [Vol. I, No. 2 (2) That the referee for next year study the ratio of nitrogen to protein in American feeding stuffs. Approved. (3) That the referee compare the various proposed methods of crude fiber determinations with the present official methods. Approved. SUGAR. It is recommended— (1) That (1) the refractometric methods for determining the moisture content of sugars, (2) the de'ermination of the moisture content of mo- lasses by means of the immersion refractometer, and (3) the direct deter- mination of sucrose in molasses, be further studied. Approved. (2) That the referee for next year study the various methods of esti- mating the copper in low grade sugar products. Approved. WATER IN FOODS. It is reeommended— (1) That comparison of drying organic or other materials at room temperature in partial vacuum and at atmospheric pressure be continued, using phosphorus pentoxid and calcium carbid as dehydrating agents. Approved. (2) That there be a further comparison of the dehydrating power of sulphuric acid, phosphorus pentoxid, calcium carbid, and metallic sodium, and any other reagent that may be found, at room temperature and at atmospheric pressure. Approved. (3) That the advisability be considered of using a general method for moisture, consisting of 24 or 48 hours’ storage over either sulphuric acid, phosphorus pentoxid, calcium carbid, or metallic sodium, at room tem- perature and at atmospheric pressure to be followed by the vacuum oven at 70°C. or 100°C. for a short time. Approved. (4) That moisture determination by the vacuum method over sulphuric acid (Bul. 122, p. 219) be made an optional official method. Approved for final action in 1914. (5) That the title of the referee on Water in Foods be changed to that of referee on Water in Foods and Feeding Stuffs. Approved. 1915] TROWBRIDGE: REPORT OF COMMITTEE B 333 ORGANIC AND INORGANIC PHOSPHORUS IN FOODS. It is recommended— (1) That methods to determine the influence of heat upon organic phos- phorus compounds in animal tissues, especially in tissues containing phos- phoric acid, be given further study. Approved. (2) That in vegetable substances (a) the completeness of the extraction, (b) the effects of using much larger amounts of magnesia mixture in the precipitation, (c) the allowing of more time for the precipitation with magnesia mixture, (d) the facilitating of the filtration by the use of the centrifuge, (e) the use of mechanical means to break up the precipitate in the acid alcohol to insure the complete solution of the phosphate, be further studied. Approved. (3) That the title of the referee on Organic and Inorganic Phosphorus in Foods be changed to that of referee on Organic and Inorganic Phos- phorus in Foods, Feeding Stuffs, and Drugs. Approved. SEPARATION OF NITROGENOUS BODIES (MEAT PROTEINS). I. Meats and beef extracts. It is recommended— (1) That the Kjeldahl-Gunning-Arnold method for determining total nitrogen in meat and beef extract be made official. Adopted, final action. (2) That in Bulletin 107, Revised, page 108, 7a, Kjeldahl-Gunning- Arnold be inserted after the word Gunning, making the sentence read: “BWmploy either the Kjeldahl, the Gunning, or the Kjeldahl-Gunning- Arnold method.” The digestion with sulphuric acid should be continued for at least 4 hours with the first two methods, and for 2 hours after the digestion has become clear with the last method. Adopted, final action. (3) That the following description of the Kjeldahl-Gunning-Arnold method be given in an appropriate place in Bulletin 107, Revised: ‘In addition to the mercury and sulphuric acid of the Kjeldahl method add 5 to 7 grams of potassium sulphate. Digest as usual but do not add any po- tassium permanganate at the end. Continue the digestion for 2 hours after the liquid has become clear or 14 hours after the digest has reached the final color.” Approved. 4 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. J, No. 2 OO Oo ITI. Nitrogenous bodies in meats and meat products. It is recommended— (1) That in Bulletin 107, Revised, page 108, 7 (b), the following method for determining insoluble and soluble protein in meat be made optional: Exhaust 7 to 25 grams of the sample (depending upon the moisture con- tent) with 330 ec. of cold (15°C.) distilled water. Make 11 successive extractions, 4 of 50 cc., 4 of 25 cc., and 3 of 10 ce. each. Dilute the ex- tract to 500 cc. and determine the total soluble nitrogen in 50 ce. De- duct the percentage of soluble nitrogen from the percentage of total nitrogen and multiply the difference by 6.25 to obtain inso'uble protein. Adopted, final action. (2) That in Bulletin 107, Revised, page 108, 7 (d) the proposed method for determining the coagulable protein in meats (see report of the referee) be made optional. Approved for final action in 1914. (83) That the Folin method for estimating creatin and creatinin in meat and beef extract be made official, the method as originally published by Folin being modified as given in the report of the referee. Approved for final action in 1914. TESTING CHEMICAL REAGENTS. It is reeommended— (1) That the lead molybdate method recommended by Krauch-Merck for the determination of molybdie anhydrid in molybdic acid, as modified by the referee, be given codperative study, together with methods for the determination of nitrogen in sodium hydroxid and impurities of ether nonvolatile at 100°C. Approved. (2) That the effect of the presence of sulphates on the accuracy of the lead molybdate method for the determination of molybdic anhydrid be studied, and that the nature and amount of the impurities in citric acid, methods for the determinations of sodium hydroxid and sodium carbonate in crude caustic soda, and the amount of phosphoric acid (P20;) in molyb- dic acid, be given further study. Approved. TANNIN, MEDICINAL PLANTS, AND DRUGS. No recommendations by the committee as no reports from referees were received. 1915] BARTLETT: REPORT OF COMMITTEE ON RESOLUTIONS 335 REPORT OF GENERAL COMMITTEE ON RECOMMENDATIONS OF REFEREES. By P. F. Trowsripce, Chairman. The committee recommends that, after any method has been adopted on first reading as provisional or official, it shall be referred to the suc- ceeding referee for collaborative work and report on the results of this collaboration before it can be finally adopted as provisional or official. Approved. The committee announces the resignation of F. W. Woll as a member of the committee on recommendations of referees and recommends the appointment of a member to take his place for the unexpired term. Approved. (R. E. Stallings, of Georgia, was appointed.) REPORT OF COMMITTEE ON RESOLUTIONS. By J. M. Barruert, Chairman. The Association of Official Agricultural Chemists cordially endorses the action of the Secretary of Agriculture in establishing a Journal of Agricul- tural Research to serve as a medium for the publication of the research work of the United States Department of Agriculture. The association will warmly welcome the enlargement of the scope of this journal so as to include the research work of the Experiment Stations and the extension of its mailing lists so as to include the names of those now included in the mailing list of the Experiment Station Record. Resolved, That the cordial thanks of this association be extended to W. D. Bige- low and A. W. Bitting for the bounteous hospitality extended and the cordial good will manifested on the evening of November 17 at the laboratories of the National Canners Association. Resolved, That the Association of Official Agricultural Chemists extend to the management of the Raleigh Hotel its sincere thanks and high appreciation for the use of the Banquet Hall and for the courtesies shown the members of this association. Resolved, That this association hereby expresses its appreciation to President Fraps for the able, impartial, and courteous manner in which he has presided over its deliberations. The report of the committee was approved. 336 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 REPORT OF THE AUDITING COMMITTEE. The following report by the secretary-treasurer was examined by the auditing committee and found to be correct: REPORT OF THE SECRETARY-TREASURER FOR THE YEAR 1912 To 1913. During the year, W. D. Bigelow, secretary-treasurer, resigned and the undersigned was appointed in his place. Separate statements are given of receipts and expenditures made by Mr. Bigelow since the report for 1912 and before his resignation and a detailed statement of the receipts and expenditures made by me since that time. On July 16, 1913, the date of the statement by Mr. Bigelow, 48 Federal and State organizations and one municipal organization had paid dues for 1912 to 1913. In addition to this, dues from two States for 1911 to 1912 are included in his report as they were received too late to be incorporated in the report for last year. The report of Mr. Bigelow as secretary- treasurer, showing a balance on hand, July 16, 1913, of $149.19, is attached. Respectfully, (Signed) C. L. ALSBERG, Secretary-Treasurer. Expenditures: July 30, 1913. 600 folders and 475 two-cent envelopes.................. $26 .97 ‘Additional: postazecncs.ot soccer ees ricer he Cec eee nee See eee eee .20 Novembertl,1913-) 200 mumberedstagster snc sec- sone saint 2.00 November 15, 1913. Dues returned to a municipal board of health...... 2.00 PotalCexpen dvb uTEss jeer cc mista cots EL ie aes sys HSL tease as een oe aT $31.17 Receipts: Received from W. D. Bigelow, balance in treasury...................+.- $149.19 Dues in stamps from South Dakota Food and Drug Commission. ....... 2.00 Total receirptsicoeaccte wes Ate tees Hares tee eesti s Se eee re eee $151.19 Balance November 15, 1913s a.con scrote sie ae oiee i esternrnis eapeuera ee ete $120.02 Respectfully submitted, (Signed) C. L. AtsBERG, Secretary-Treasurer. Examined and found correct: (Signed) JoHN PHitiirs STREET, (Signed) W. H. McIntire, (Signed) H. D. Haskins, Auditing Committee. The association adjourned at 12.40 to reassemble at 2 o’clock. WEDNESDAY—AFTERNOON SESSION. A report on medicinal plants and drugs was read by the referee, H. A Seil, but not submitted for publication in the Proceedings. REPORT ON SYNTHETIC PRODUCTS. By W. O. Emmury, Associate Referee. During the past year interest in coéperative work on drug products has been unusually marked, in that the number of those indicating a de- sire to participate in the work has been much larger than in any previous year. The samples examined, bearing the numbers 13, 14 and 15, were derived from tablets purchased in the open market and alleged to have the following active constituents: 13: ‘“Acetanilid 2 grs. Quinin sulf. 1 gr. in each tablet;” 14: ‘Acetphenetidin 2.5 grs. Quinin sulf. 2.5 grs. in each tablet;” 15: “ Acetanilid 2 grs. Quinin sulf. 2 grs. Morph. sulf. 4 gr. in each tablet.’”’ The tablets, the average weights of which had been pre- viously determined, were powdered, made up into suitable samples and submitted along with the following methods of procedure to the co‘ perat- ing chemists: ESTIMATION OF ACETANILID AND QUININ SULPHATE. The separation of these two substances is based on the fact that the bi- sulphate of quinin in aqueous-acid solution is practically insoluble in U. 8. P. chloroform, while acetanilid under the same conditions is readily taken up by this solvent. The procedure, therefore, resolves itself into the following steps: Ascertain the weight of 20 or more tablets, reduce them to a powder and transfer to a glass-stoppered or well-corked flask. Weigh out on a metal scoop, watch glass, or other convenient object an amount of the powdered sample equal to or a multiple of the average weight of one tablet, transfer to a separatory funnel (Squibb form), add 50 ce. of chloroform, 20 cc. of water and 10 drops of dilute sulphuric acid, sufficient at least to insure a slight excess of this reagent in the mixture. Shake for some time vigorously, allow to clear, then draw off the solvent through a small pledget of cotton and a small (5.5 cm.) dry filter into a 200 cc. Erlenmeyer. Repeat extraction twice, using the same amount of chloroform as in the first operation. Use a fresh pledget of cotton for each withdrawal of solvent, putting the moist cotton after passage of the chloroform into the filter, where with the latter it is allowed to dry spontaneously, or by placing a ‘ew moments on the cover of a steam bath. On completion of the third extraction the separation of the two ingredients 337 338 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMIsTs [Vol. I, No. 2 in question is practically complete, all the acetanilid being in the chloroform, while the quinin remains in the aqueous-acid solution, with traces also in both cotton and filter. Acetanilid.—Distill the chloroform from the three extractions by the aid of gentle heat down to about 10 cc., add 10 ce. of dilute sulphuric acid (1:10 by volume), continuing the distillation until all the solvent has passed over. Remove to a steam or vapor bath and digest for about one hour or until the liquid has evaporated to about two-thirds the original volume. Add 20 cc. of water, digest 30 minutes longer, add 10 ec. of concentrated hydrochloric acid, then titrate with a standard solution of potassium-bromid-bromate, substantially as outlined in the method for the estimation of acetanilid and caffein. Quinin sulphate.—Wash filter and cotton used in drying the chloroform solution of acetanilid once with 5 cc. of water, allowing latter to run into the aqueous-acid solution of quinin. Add solid sodium bicarbonate (or aqueous ammonia) in slight excess, then extract with three 50 ec. portions of chloroform, washing each portion in rotation with 5 ec. of water, and passing the solvent after clearing through cotton and a dry filter, exactly as in the extraction of acetanilid. Distill the chloroform from the several extractions down to about 10 cc., then, if it seems desirable to weigh the quinin as such, transfer residue to a small tared beaker by pouring and subse- quent washing with chloroform, evaporate to apparent dryness on the steam bath, heat for an hour at 125°C. in an air bath, cool, and weigh. If, as is usually the case in combinations like the one at present under examination, the weight of quinin sulphate is desired, distill the chloroformic solution of quinin to apparent dryness by means of gentle heat, dissolve residue in 3 to5cc. of neutral alcohol (just sufficient to prevent precipitation by the standard acid) and titrate with fiftieth-normal sulphuric acid (using two drops of methyl red solution as indicator) until the color changes to faint red. Remove to steam bath and heat until most of the alcohol has been ex- pelled, the color of the liquid having in the meantime become yellow again. Now add sufficient acid to restore the faint red coloration, note number of cubic centi- meters expended, then multiply same by 8.66, the value of 1 cc. of fiftieth-normal sulphuric acid in milligrams of quinin sulphate (C2H24N202)27H2S047H:O, to get the weight of this substance in sample taken. In the event that the quinin as such has first been weighed, the weight should be further checked by titration, substan- tially as outlined. Comments and suggestions —The composition of many medicinal prepa- rations in pill or tablet form is frequently of such a nature as to com- pletely inhibit any rational separation of the active organic constituents by means of immiscible solvents in the ordinary separatory funnel, owing to the formation of persistent emulsions even on cautious agitation. To obviate this difficulty various flattened types of separators have been suggested by workers in the Bureau of Chemistry, one of which (illus- trated in J. Amer. Chem. Soc., 1913, 85: 295) yields gratifying results. A 20-minute treatment on the rotating table suffices to effect a maximum distribution of the substances involved. Wherever available such separa- tors will be found highly advantageous in obviating the delay and annoy- ance occasioned by emulsifying mixtures. Since quinin sulphate readily loses a portion of its water of crystallization when exposed to dry air, the amount of sulphate found, whether calculated from the quinin as 1915] EMERY: SYNTHETIC PRODUCTS ; 339 weighed or from that determined volumetrically, need not necessarily correspond with the declared amount of this commodity indicated on the sample under investigation. As far as the separation is concerned the method differs in no particular from that outlined for the preceding com- bination. In the estimation, however, the procedure as applied to acet- phenetidin is materially shortened, in that the physical properties of this substance permits its being weighed directly. Acetphenetidin.—Extract with three 50 ce. portions of chloroform, washing each portion in rotation in a second separator (Squibb form) with 5 ec. of water and passing solvent after clearing through a pledget of cot- ton and small dry filter into a 200 ec. Erlenmeyer. Distill the chloro- form from the combined extractions down to about 10 ec., transfer resi- due by pouring and washing with additional chloroform to a small tared crystallizing dish, allow to evaporate spontaneously or at a moderate heat on the steam bath, cool, and weigh at intervals until the loss does not exceed 0.5 mg. Quinin sulphate-—Follow directions as given under acetanilid and quinin sulphate. Comments and suggestions.—In all cases where cotton is used to remove suspended moisture from chloroform as suggested in the foregoing work, the material is inserted in the outlet tube, thus intercepting most of the moisture and any suspended matter that might otherwise clog the filter. ESTIMATION OF ACETANILID, QUININ SULPHATE AND MORPHIN SULPHATE. As in a mixture of acetanilid and quinin sulphate, so likewise in the present combination, the alkaloidal constituents in aqueous-acid solu- tion are separated from the third by virtue of the insolubility of their sulphates in chloroform. The separation of the alkaloids themselves is based on the ability of morphin to yield with an alkaline base a morphinate insoluble in chloroform. The procedure thus resolves itself into the following particulars: Acetanilid.—Transfer to a separatory funnel an amount of the powdered sample equal to or a multiple of the average weight of one tablet (the amount of morphin sulphate represented should not be less than 0.25 grain), add 20 ec. of water and 10 drops of dilute sulphuric acid, then extract with three 50 cc. portions of alcohol-free chloroform, the subsequent manipulations being substantially as directed for this ingredient in the combination: Acetanilid and quinin sulphate. Quinin sulphate.—Wash filter and cotton used in drying the chloroformic solution of acetanilid once with 5 ce. of water, uniting latter with the aqueous-acid solution of quinin and morphin. To this solution add 4 to 5 ec. of aqueous sodium hydroxid (5 grams of pure sodium hydroxid in 50 ce. of water), then extract with three 50 cc. and one 25 ce. portions of U. S. P. chloroform, transferring latter in rotation to a second separator (Squibb type) containing 5 ec. of water, wash and pass the nearly clear chloroform through a pledget of cotton and small filter into a 200 ec. Erlen- 340 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 meyer, from which the solvent is later removed by gentle distillation and the residue titrated, substantially as outlined for quinin in the combination: Acetanilid and quinin sulphate. Morphin sulphate.—Wash filter and cotton employed in the preceding operation with 5 ee. of water, which latter together with that portion used to wash the chloro- formic solution of quinin is united with the aqueous-alkaline solution of sodium morphinate. Now add 0.5 gram of ammonium chlorid (an amount slightly in excess of that required to free the morphin as well as convert the sodium hydroxid into sodium chlorid) and to the resulting ammoniacal solution add 45 ee. of U. S. P. chloroform and 5 ec. of alcohol, then extract and draw off solvent into a second separator containing 5 ec. of water, wash, allow to clear, then pass chloroform through cotton and filter into a 200 ec. Erlenmeyer. Repeat extraction and subse- quent washing with one 50 cc., one 40 cc. and one 30 ce. of U.S. P. chloroform, finally collecting the solvent from all extractions in and distilling from the aforesaid Erlenmeyer down to about 10 cc. Transfer by pouring and washing with additional chloroform to a small tared beaker, evaporate on the steam bath to dryness, cool and weigh the residual morphin now appearing as a transparent varnish. To render erystalline, dissolve by warming with about 1 ce. of neutral alcohol, and about the same quantity of water drop by drop, rubbing the glass with a rod to induce erystal- lization, then evaporate slowly on the steam bath to dryness. Cool and weigh a second time. Check the weight of morphin thus found by titration with fiftieth- normal sulphuric acid, using a drop of methyl red solution as indicator. To this end dissolve the morphin in 1 to 2 ce. of warm neutral alcohol; then after solution is complete add the acid till the color changes to faint red. Evaporate most of the alcohol on the steam bath, and in the event that the color has reverted to yellow add just sufficient acid to restore the faint red coloration. Note volume of acid expended, then multiply the number of cubic centimeters by 7.53 (the number of mil- ligrams of morphin sulphate equivalent to 1 cc. of fiftieth-normal sulphuric acid) to ascertain the quantity of morphin sulphate in the sample taken. The amount of anhydrous or of crystallized alkaloid can also be determined from the titration value by means of the proper factor. Comments and suggestions—For the purpose in question alcohol-free chloroform may be conveniently prepared by washing the pharmacopoeial product several times with water. All cotton used for drying chloroform should first be freed from fatty material of other extractives by thorough washing with this solvent, all of which latter may be regained by distilla- ‘tion. In the various operations involving fixation and subsequent libera- tion of morphin by means of fixed alkali and ammonium chlorid, strict attention should be paid to the matter of adding these reagents, since any undue excess of either might nullify the entire procedure. Any large ex- cess of sodium hydroxid would naturally require for its reduction a cor- respondingly large amount of ammonium chlorid, the latter in turn yield- ing its prorata of ammonium hydroxid, relative large quantities of which through interaction with sodium chlorid tend to inhibit any permanent liberation of the alkaloid and thus prevent a complete extraction. Fur- thermore, the presence of relatively large quantities of ammonium chlorid as such operates to a partial retention of morphin in solution, due in part, 1915| EMERY: SYNTHETIC PRODUCTS 341 possibly, to the formation of the hydrochlorid of this alkaloid. In spite of all precautions in the matter of excluding impurities from the morphin, the amount of this substance as found by weight will usually be greater than that determined volumetrically. In order to insure greater accuracy in volumetric operations with alkaloidal substances, as quinin and mor- phin, the suggestion is made, in all cases where possible, that the strength of the standard acid used be checked by titration against the pure alkaloid under examination. COOPERATIVE RESULTS. Coéperative results on synthetic products. 131 142 158 2 QUININ SUL- QUININ SUL- QUININ SUL- | MORPHIN SUL- ANALYST PHATE ACET- PHATE PHATE PHATE ACET- a es | ACEI esse Gravi- | Volu- | Tiptn | Gravi- | Volu- ANILID | Gravi-| Volu- | Gravi- Volu- metric | metric metric | metric metric | metric | metric | metric grains | grains | grains | grains | grains | grains | grains | grains | grains | grains | grains = 1.90 0.98 0.97 2.51 2.47 2.38 1,92 1.94 1.89 0.109 0.093 UL, A Brown, Lexington, Ky...) 41/29 | 9.99 | 0.97 | 2.54 | 2.45 | 2.41 | 1.91 | 1.95 | 1.89 | 0.105] 0.090 J.¥. Darling New York, N. Y 1.96 1.01 1.02 2.51 2.54 2.47 1.93 1.86 1.82 0.184 0.162 . pate fl 92 0.98 0.87 2.47 2.10 2.01 1.90 1.87 IPA || agape || sages iSe Ob every SERINE AE A aa | clan |) en | ace I aa |) amy || fem | ea I age |) cca ll conn W. O. Emery, Washington, D. . (Ch sociage Ane ae DOG Cems |e ee) ) Hone 0.96 2.49 36 2.45 1.90 a58C 1.91 0.080 0.068 H. Engelhardt and O. E, Win- fi 91 0.98 0.94 2.43 2.22 2.21 1.97 1.83 1.81 0.200 0.148 ters, Baltimore, Md.......... \1 98 1.04 0.95 2.56 2.42 2.36 1.94 1.86 1.79 0.208 0.151 H. C. Fuller, Washington, D.C 1.96 1.00 oe 2.49 2.40 Rae 1.89 1.96 Arie exticcne 0.080 - 2.01 0.96 2.43 2.30 2.05 UE ODS etetetst'« 0.128 c 8 Jashs.... fo (Gee Si ay SEO ES i GOW shoe NV Oa eeeb ll pee eal shes ED |] oben 0.093 s - r. fi 98 1.01 0.93 2.49 2.42 2.35 1.93 1.94 1.93 0.138 0.078 E. G. Grab, Nashville, Tenn...-1)1 9g | 0.98 | 0.96 | 2.49 | 2.42 | 2.29 | 1.92 | 1.93 | 1.94 | 0.133] 0.089 C. B. Morison, New Haven, (CATO. 3 5: Bono BEeone CeEEE Coes 1.99 Anec 0.93 2.47 ange 2.28 1.99 aire SY IP aaaae 0.090 E. E. Sawyer, Orono, Me.®.......| 2.03 0.99 0.93 2.64 2.48 2.26 1.83 1.90 1.64 0.129 0.068 H. A. Seil, New York, N. Y..... 1.98 1.00 1.00 2.44 2.53 2.44 1.90 2.02 TEC) co adan 0.132 O. Stockinger, Philadelphia, f1.97 0.98 0.95 2.51 2.47 2.30 1.92 1.94 0.043 | 0.033 LEA necioctadisio s do casasneeedenaed \ 1.96 308 2.48 2.45 2.30 1.92 erates 1.93 0.033 0.025 E. R. Tobey, Orono, Me.®....... 2.00 0.91 0.89 2.53 2.44 2.37 1.89 1.91 1.81 0.090 0.076 A. R. Todd, Lansing, Mich...... 1.90 ate 0.96 2.42 So 2.45 1.37 asta IP | Specac 0.500 C. C. Wright, Washington, D. f2.00 1.00 0.92 2.50 2.46 2.42 1.91 1.87 1.80 0.129 0.098 Me eet cay eisie's -'s)e aisiesicerssa 2.00 1.01 0.95 2.50 2.47 2.3! 1.90 1.88 1.79 0.125 0.101 Wtealarad antec: icin mrse'teiis 3 2.00 1.00 2.50 2.50 2.00 2.00 0.125 1 Average weight of the single tablet, 0.2632 gram. 2 Average weight of the single tablet, 0.3851 gram. 3 Average weight of the single tablet, 0.2939 gram. 4 Reported by C. W. Johnson. 5 Reported by J. P. Street. » Reported by J. M. Bartlett. 7 Reported by C. I. Vanderkleed. Examination of the data herein presented by 15 chemists indicates that little or no trouble was experienced in estimating acetanilid and acet- 342 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 2 phenetidin, the recoveries corresponding very closely with the amounts of active ingredients alleged to be present. Some of the workers appear to have had difficulty with quinin, while in the separation of this substance from morphin and the estimation of this latter constituent considerable trouble must have been experienced, as indicated by the widely-varying results, and by the comments of the workers themselves. In manipulat- ing the several mixtures with the ordinary type of separatory funnel, more or less difficulty would necessarily be encountered owing to a tend- ency of the tablet excipients to favor emulsification. No.15 in particular was undoubtedly the most difficult problem in this respect, hence it is not a little to be wondered at that some of the least experienced workers were able to effect a gross separation at all with the apparatus generally available and to arrive at the commendable results presented. Appar- ently one-third of the workers had never before had opportunity to examine such samples. About half those coéperating had no suggestions or criti- cisms to offer relative to the methods followed, from which the inference is drawn that little or no trouble was encountered by them in carrying out the various operations. Others, however, were kind enough to formu- late their experience in the following terms: COMMENTS BY COLLABORATORS. L. A. Brown had no criticism or suggestion except as to Sample 15, in which case it was necessary to use more than one tablet in order to get a workable amount of morphin. He found 3 to 50 ce. portions of chloroform insufficient to extract all the acetanilid. Furthermore, in the titration of quinin, from a sample of three tablets used by him, the alkaloid was dissolved in 50 ce. of neutral alcohol and then aliquot portions titrated. J. F. Darling believed the method would be improved by a provision for removal of any insoluble gum prior to extraction. His separation of quinin from morphin was not sharp, some of the former appearing with the morphin residue. E. O. Eaton’s experience with No. 15 was so unsatisfactory that no analytical returns were made on this mixture. No other criticisms reported. One would infer, however, that some difficulty had been encountered with quinin in No. 14. H. Engelhardt and O. E. Winters found the method satisfactory except as affected by persistent emulsification in No. 15. £.G.Grabreferred in connection with the hydrolysis of acetanilid and acetphenet- idin to the transfer of the 10 cc. of chloroformic residue containing one of these substances after addition of dilute sulphuric acid directly to steam bath, as being preferable to further recovery of solvent, since the last runnings of chloroform are likely to carry suspended water and thus contaminate the main portion. His point is that little is gained in attempting to recover the chloroform after addition of acid. C. E. Vanderkleed is of the opinion that accurate determinations of morphin sulphate are scarcely possible when working on such small quantities of material as called for in the method. The referee believes notwithstanding that the method will yield accurate results with an amount of morphin sulphate as small as one-fifth of a 1915] ST. JOHN: ANALYSIS OF MEDICATED SOFT DRINKS 343 grain, provided the directions are followed closely as to detail and emul- sification is avoided by the use of the so-called ‘‘terrapin” form of sepa- rator. This belief is based on results obtained with controls. H. A. Seil, in commenting on the titration of quinin stated that solution of the alkaloid in standard fiftieth-normal acid with subsequent titration of excess acid with fiftieth-normal alkali gives more accurate results and a sharper end point. Methyl Red as indicator appears to be better adapted to the alkaline end point than to the acid end point. In the separation of acetanilid and acetphenetidin from quinin, a wash with dilute acid in the second separatory funnel insures a cleaner separation. This method would do away with filtration through cotton plugs and would simplify the method from a mechanical standpoint. Similarly, a wash with dilute alkali in the second separatory funnel would insure a cleaner separation in the estimation of quinin from morphin. - CONCLUSION. Taken as a whole, the results are very promising and indicate that the methods are correct in principle, requiring slight changes, if any, in the matter of detail. A paper on the Estimation and Separation of Antipyrin from Various Synthetic Products by Means of its Periodid, by W. O. Emery and 8. Palkin, was read by Mr. Palkin. It has since been published under the title: Studies in Synthetic Drug Analysis. JI—Estimation of Antipyrin, in the Journal of Industrial and Engineering Chemistry, 1914, volume 6, page 751. The associate referee on medicated soft drinks had no report but recom- mended that the following paper by Mr. St. John be read. SUGGESTIONS ON THE ANALYSIS OF MEDICATED SOFT DRINKS. By B. H. Sr. Joun. SALICYLIC ACID, BENZOIC ACID, AND SACCHARIN. Some work has been done toward finding a satisfactory method for the separation and accurate determination of salicylic and benzoic acids and saccharin, and as a result of this work the following method has been evolved which will permit also the determination of caffein in the sample if desired. Method. Make 100 cc. of the sample distinetly acid with sulphuric acid and shake with 50 ce. of a mixture of 7 parts of chloroform and 3 parts of alcohol. Run the chloroform layer into a second separator and wash with a strong solution of sodium carbonate. If it is desired to determine caffein, wash the chloroform, after it has 344 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMIsTS [Vol. J, No. 2 been shaken with a sodium carbonate solution, with water in a third separator and filter off through a filter wetted with chloroform. Repeat this operation with suc- cessive portions of 50 cc. of the chloroform-alcohol mixture. It will be found that the caffein will be completely extracted and that the caffein residue obtained by evaporating this washed chloroform will be almost free from coloring matter, as a large proportion will be removed by the sodium carbonate. The sodium carbonate solution will contain benzoic and salicylic acids and saccharin if present. The salicylic acid can be readily separated and determined by adding iodin solution in excess, warming on the steam bath for an hour or so, and filtering off the rose-colored precipitate whose composition is (CgH2I.0)2, drying and weighing. The weight of the precipitate times 0.4657 gives the weight of sodium salicylate present. When the filtrate is acidified and the iodin removed by hypo- sulphite or sulphurous acid the benzoic acid and saccharin can be shaken out with the above mixture of alcohol and chloroform. The separation of benzoic acid and saccharin is more difficult, as under the conditions where benzoic acid is distilled with steam some saccharin will come over also. It is possible that the saccharin under proper condi- tions could be precipitated almost completely as silver salt, at least to such an extent that none would distill off with steam with the benzoic acid; or the benzoic acid may be sublimed off from the saccharin by being heated in an oven to constant weight, the residue being saccharin. The saccharin can also be separated by transforming it to salicylic acid by heating with saturated solution of sodium hydroxid, and precipitating the salicylic acid with iodin as just described. The quickest and easiest method is perhaps the sublimation method. PHOSPHORIC ACID. The present method for the determination of phosphoric acid requires much time for making the precipitation, which even then is not always complete in such preparations as medicated soft drinks. Furthermore, the composition of the precipitate frequently is not that of magnesium am- monium phosphate or phosphomolybdic acid, in which form it is desired that phosphoric acid be separated. The following modification is the result of some work done in the past year on the subject, and it is desired to offer this with whatever suggestions may be made towards its improvement for céoperative work during the following year. Method.—Dilute the sample, usually 25 to 100 ce., with water, add 30 cc. of aqua ammonia, surround the beaker with ice, and run in slowly from a burette 50 ce. of magnesia mixture, stirring vigorously. Continue the stirring, preferably, of course, by means of a stirring machine, for a half hour, when precipitation will be found to be complete and the precipitate in the form of magnesium ammonium phosphate as desired. Filter off the precipitate and wash with a mixture of 1 volume of aqua ammonia and 3 volumes of water (which should be as cold as possible) until the washings have no appreciable color. Dissolve in as small as possible a quantity 1915] ST. JOHN: ANALYSIS OF MEDICATED SOFT DRINKS 345 of warm dilute nitric acid, then add about 50 to 75 cc. of ammonium molybdate so- lution and allow to stand 12 hours in a warm place at 50° to 60°C. This time may be materially shortened by stirring. Filter and wash with a solution containing 90 grams of ammonium nitrate and 200 cc. of nitric acid (specific gravity 1.2) per liter. Dissolve the precipitate of phosphomolybdic acid in aqua ammonia, make up to approximately 100 cc., and then proceed as in the first precipitation of the mag- nesium ammonium phosphate, ignite the precipitate after filtering and washing, and weigh the magnesium pyrophosphate. GLYCERIN. In the determination of glycerin in medicinal soft drinks, if a sample of 10 to 25 ce. is taken, the method given in Bulletin 107, Revised, for the determination of glycerin in wines will give a quite satisfactory separation of glycerin from the sugars and gums present. Caffein and other sub- stances are likely to be present, however, in the residue from this method, and consequently its weight does not represent glycerin alone. It is advisable to dissolve the residue in water and oxidize the glycerin with alkaline permanganate to oxalic acid, removing the excess of permanganate with c.p.methyl alcohol. Make up solution to a suitable volume, filter, acidify an aliquot taken from the filtrate, boil off the carbon dioxid, make alkaline with ammonium hydroxid and precipitate with calcium chlorid solution. The calcium oxalate can be ignited and weighed as calcium oxid or calcium sulphate. or dissolved in sulphuric acid and titrated with permanganate. The association adjourned. OFFICERS AND REFEREES OF THE ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS, 1913-1914 Honorary President. H. W. Witey, Washington, D. C. President. E. F. Lapp, Fargo, S. Dak. Vice-President. C. H. Jones, Burlington, Vt. Secretary. C. L. Auspere, Washington, D. C. Additional Members of the Executive Committee. J. D. Turner, Lexington, Ky. W. F. Hanp, Agricultural College, Miss. Referees. Phosphoric acid: A. J. Patten, East Lansing, Mich. Nitrogen: Determination: R. N. Brackett, Clemson College, 8. C. Separation of nitrogeneous substances: A. W. Bosworth, Geneva, N. Y. Potash: E. E. Vanatta, Columbia, Mo. Soils: J. W. Ames, Wooster, Ohio. Dairy products: L. I. Nurenberg, Boston, Mass. Feeds and feeding stuffs: G. L. Bidwell, Washington, D. C. Food adulteration: Julius Hortvet, St. Paul, Minn. Sugar: No appointment. Tannin: C. B. Bacon, Washington, D. C. Insecticides: R. C. Roark, Washington, D. C. Inorganic plant constituents: B. E. Curry, Durham, N. H. Medicinal plants and drugs: H. A. Seil, Pittsburgh, Pa. Water: W. W. Skinner, Washington, D. C. Testing chemical reagents: J. B. Rather, College Station, Tex. Water in foods and feeding stuffs: W. J. McGee, Washington, D. C. Organic and inorganic phosphorus in foods, feeding stuffs, and drugs: E. B. Forbes, Wooster, Ohio. Associate referees. Phosphoric acid: L. 8. Walker, Amherst, Mass. Nitrogen: Determination: H. D. Haskins, Amherst, Mass. Special study of Kjeldahl method: I. K. Phelps, Washington, D. C. Separation of nitrogeneous substances: Milk and cheese: Leroy S. Palmer, Columbia, Mo. Vegetable proteins: T. B. Osborne, New Haven, Conn. Meat proteins: A. D. Emmett, Urbana, III. 346 19165] OFFICERS, REFEREES AND COMMITTEES 347 Potash: Determination: T. D. Jarrell, College Park, Md. Availability: E. E. Vanatta, Columbia, Mo. Soils: A. W. Blair, New Brunswick, N. J. Nitrogeneous compounds: C. B. Lipman, Berkeley, Calif. Alkali in soils: R. F. Hare, State College, N. Mex. Dairy products: W. M. Clark, Washington, D. C. Feeds and feeding stuffs: A. C. Summers, Columbia, S. C. Feed adulteration: Carlton Cutler, W. Lafayette, Ind. Food adulteration: Colors: W. E. Mathewson, New York, N. Y. Saccharine products: F. L. Shannon, Lansing, Mich. Fruit products: H. C. Gore, Washington, D. C. Wine: B. G. Hartmann, Chicago, III. Beer: J. G. Riley, Boston, Mass. Distilled liquors: A. B. Adams, Washington, D. C. Vinegar: E. H. Goodnow, St. Paul, Minn. Flavoring extracts: A. E. Paul, Chicago, III. Spices: R. W. Hilts, San Francisco, Calif. Baking powder: R. E. Remington, Fargo, N. Dak. Meat and fish: H. S. Grindley, Urbana, III. Fats and oils: R. H. Kerr, Washington, D. C. Dairy products: Julius Hortvet, St. Paul, Minn. Cereal products: B. R. Jacobs, Jersey City, N. J. Vegetables: E. W. Magruder, Richmond, Va. Cocoa and cocoa products: H. C. Lythgoe, Boston, Mass. Tea and coffee: J. M. Bartlett, Orono, Me. Preservatives: A. F. Seeker, New York, N. Y. Heavy metals in foods: E. L. P. Treuthardt, Washington, D. C, Sugar: M. N. Straughn, Washington, D. C. Insecticides: O. B. Winter, East Lansing, Mich. Medicinal plants and drugs: Synthetic products: W. O. Emery, Washington, D. C. Medicated soft drinks: W. F. Sudro, Fargo, N. Dak. Medicinal plants: Frank Rabak, Washington, D. C. Alkaloids: H. Buchbinder, Washington, D. C. Balsams and gum resins: E. C. Merrill, Washington, D. C. Water: H. P. Corson, Urbana, III. Tannin: F. P. Veitch, Washington, D. C. Inorganic plant constituents: R. W. Thatcher, St. Paul, Minn. Testing chemical reagents: R. C. Wells, Washington, D. C. PERMANENT COMMITTEES. Food Standards. William Frear, State College, Pa., chairman. E. H. Jenkins, New Haven, Conn. R. E. Doolittle, New York, N. Y. B. B. Ross, Auburn, Ala. H. E. Barnard, Indianapolis, Ind. 348 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMIstTs [Vol. I, No. 2 Co6PERATION WITH OTHER COMMITTEES ON Foop DEFINITIONS. William Frear, State College, Pa. Julius Hortvet, St. Paul, Minn. J. P. Street, New Haven, Conn. Recommendations of Referees and Revision of Methods. (Figures in parenthesis refer to year in which appointment expires.) P. F. Trowbridge, chairman. SupcomMitTEE A: A. J. Patten (1918), W. W. Skinner (1916), B. B. Ross (1914), chairman, Alabama Polytechnic Institute, Auburn, Ala. (Phosphoric acid, nitrogen, potash, soils, inorganic plant constituents, insecticides, water.) Suscommirrere B: R. E. Stallings (1918), P. F. Trowbridge (1916), HZ. M. Chace, (1914), chairman, Bureau of Chemistry, Washington, D. C. (Dairy products, foods and feeding stuffs, sugar, water in foods and feeding stuffs, organic and inorganic phosphorus in foods, feeding stuffs, and drugs, separation of nitrog- eneous bodies, testing chemical reagents, tannin, medicinal plants and drugs.) Suscommittee C: L. M. Tolman (1918), H. ZH. Barnard (1916), chairman, Indianapo- lis, Ind., C. D. Howard (1914). (Food adulteration.) SpeciaL ComMUrreEs. Editing Methods of Analysis (Bulletin 107, Revised). J. K. Haywood, Washington, D. C., chairman. W. A. Withers, Raleigh, N. C. J. P. Street, New Haven, Conn. . A. F. Seeker, New York, N. Y. G. W. Hoover, Chicago, Ill. B. L. Hartwell, Kingston, R. I. Proposed Journal of Agricultural Chemistry. E. F. Ladd, Fargo, N. Dak., chairman, C. H. Jones, Burlington, Vt. C. L. Alsberg, Washington, D. C. J. D. Turner, Lexington, Ky. W. F. Hand, Agricultural College, Miss. Study of Vegetable Proteins. T. B. Osborne, New Haven, Conn., chairman. (Two additional members to be appointed by chairman.) Availability of Phosphoric Acid in Basic Slag. C. B. Williams, West Raleigh, N. C., chairman. C. G. Hopkins, Urbana, Ill. H. D. Haskins, Amherst, Mass. B. L. Hartwell, Kingston, R. I. J. A. Bizzell, Ithaca, N. Y. Review of the Analysis of Lime Sulphur Solutions. G. A. Hulett, Princeton, N. J., chairman. C. 8S. Cathcart, New Brunswick, N. J. H. H. Hanson, Orono, Me. INDEX TO VOLUME I, NUMBER 2 Antipyrin, estimation, paper by Emery and Palkin, reference................. 343 PUG IN ee COMMItLEe BLE DOLbaeiasqe-tas meine cists ae eee ee eee ee 336 ECOL NEMO) aaa eee aate SBM omede Meme res dab Cra ce pn onoed aden hades 329 Balle yeerepoctronud alrys PLOdUCUS mer acectciisaci ciccics eae cree eee eee 289 Baking powders, lead content, paper by Seeker and Clayton.................. 264 Bartlett, report by committee on resolutions....................00.ceeeeeceees 335 TEPOLtOMMLe ay AN) COMES pa eiecsreess he eke, op ascierasc Ves atosera els erent Eo: 203 Beer recommendations by, Committee Gx. oa... 6.2 vce ck sn ores susan teeoceee eee 283 Biesterfeld, supplemental report on dairy products (adulteration)............. 194 (Cannedstoods reportibys Viaerudermeee seed erieeeiaccieeee eee nee 199 Casein, precipitation, paper by Van Slyke and Winter........................ 281 Cereal products, recommendations by Committee C...................2-.0000- 286 LECOMMeNnGALLONS| Dye WHEN eer Acros eeeenitice mer 199 FEDOLUEL Dy) WiRLbE Nerja teaser Cantal pees le ee ee 195 hocolatesmilke reportibyslytheoesecneaemeen eee soe eee canons aes 200 Clayton and Seeker, paper on lead in baking powders....................0000- 264 Cocoa and cocoa products, recommendations by Committee C................. 286 recommendations by Lythgoe...................... 202 IeloKod eh Loni? WAAAY goounonosaoboacoosmaossauuac es 200 Coffee and tea, recommendations by Bartlett.....................0.0eeee scene 208 recommendations by; Committee Cn. s5 season ee eee 287 TEPOLLD Ys BATtle tte jae eves. Cis vies a Aer eRe rch Pte e 203 (Colors) recommendation by, Committee! ©..2... 24-422. cngnenees see ee ee eee ser 282 Committee B on recommendations of referees, report.............. 000... 0000. 331 C on recommendations of referees, report......................... 282 @oukwwelycerinvinimestiextractes qas-mecienc near ee aera eee ne 279 Crossereportonisugar and molassessscni-cac tence ne eee nee eaeee 314 Dairy products, apparatus for analysis, talk by Hand, reference.............. 195 lime as neutralizer, paper by Wichmann, reference............ 195 FECOMMENG aon yaball eyes Peeeeee ence einen ner 289 TECOMMend Aion Dye aAtricka 2 eee Meese eres ees 289 recommendations by Committee B........................:.. 331 TOPOLE, Py gall C Vp 5: caj- fs siereis se) evaueseyadepeyees lakers edtvouersiee oe ea ers 289 Dairy products (adulteration), recommendations by Committee C............. 286 recommendations by Hortvet.................. 194 report. by, HOntvets ses okeat an cairo 186 supplemental report by Biesterfeld............ 194 Davidson, report by committee on nominations..:................00cee cesses 288 Distilled liquors, recommendations by Committee C.......................... 283 Drugs and medicinal plants, report by Seil, reference.....................00. 337 349 350 INDEX merry, report on synthetic) products... eine e es) aria eee eee te 337 and Palkin, paper on estimation of antipyrin, reference...... strana Odo Emmett, report on separation of nitrogenous bodies (meat proteins).......... 267 Exner paper on the sublimator sas aee ee ter ae ioe eres tee eres 208 Fats and oils, reeommendations by Committee C............................. 285 Tecommendationss Dye Chie ee ee ee eee eee eee 186 report by Woerr: s.varsis cise secs eal ecopeeete ee eeatsl hovers stereo re Seasees ree 181 Feeds and feeding stuffs, recommendations by Committee B.................. 331 recommends tions by, JONES.) hes oe eee ee 312 TEPOLUIby: JONES sito icicfe ese vers ene sise tipo eos niece 289 Flavoring extracts, recommendations by Committee C........................ 284 Foods, heavy metals content, recommendations by Committee C.............. 287 recommendations by Loomis..............-..... 254 TEPOLtb yy, HOOMISHs chee aes eee eee 244 supplementary report by Treuthardt............ 254 phosphorus content, recommendations by Committee B................ 338 report by Forbes and Wussow................-..- 221 tinicontent; report! by, Mreuthardte-... 3. eee ee eee eee 254 water content, recommendations by Committee B..................... 332 recommendations by McGee....................2000205: 218 report»by McGees. oi dsyecte hs cee in nee ae eee 218 Forbes and Wussow, report on inorganic phosphorus in foods................. 221 Fruit products, recommendations by Committee C..........................5- 282 Glycerin;in-meat extracts, paper by Gooks--.-5-- 6 --4-e teen e eee ee eee eee 279 Hand, talk on apparatus for analysis of dairy products, reference............. 195 Hortvet, report on dairy products (adulteration) ........ fra gageraka aucterekesee een 186 Jones report on feedsiand feedimeasvulisermre ee sees ae ie ee ee eee 289 Journal of Agricultural Research, endorsement:...-. +2222 see eee eens 335 Kerns reportionitatsandsonlsen- ee rteer eee ee peed OPE eid See israel ee 181 Lead, in baking powders, paper by Seeker and Clayton....................... 264 Lime, neutralizer in dairy products, paper by Wichmann, reference........... 195 Loomis} :report/onheavy;metals ini foods!--.4.-sec- eee eee eae 244 Lythgoe, report on cocoa and cocoa products................--:.-2s++es--ee-- 200 McGee; reportionnwater in toodse. 2) see ena. ccna r ere nee Pee 218 Magruder, report on vegetables. . mae Sores sist sod ees ysis a e/a oleic eee 199 Meat and fish, recommendations by Committee c. ih Ga eee Pansere ne ope re eee eee 285 recommendationsiby smiuhee eee ere eee eee ere 180 report by Smith.. tN x cece Wad acete sio ee eee CO 170 extracts, glycerin content, paper by Cook! Piadsaecgiagainilsrie a/c eee eee 279 proteins, separation, recommendations by Committee B................. 333 recommendations by Emmett...................... 278 TEPOLt Ye HMMebie cece cere ere ete eee eee 267 Medicated soft drinks: srepontibyists J ObL sere eee ee eee eee eee 343 Medicinal plants and drugs, report by Seil, reference................-......-. 337 INDEX 351 Neeting places; Invitiationay. cai cisk crescents ee las bse, oeibcovald 330 Metals, heavy, in foods, recommendations by Committee C................... 287 recommendations: by Woomisse-eus-.:.<.---+--..0.-+- 254 reportiby Loomis:.)-: ease tete tens say sia $ 244 supplementary report by Treuthardt................. 254 Molasses and sugar, recommendations by Committee B....................... 332 recommendations by, Grossse--ee eee Een eee eee ene see 317 TEPOLt iby: CLOSSHe km: sia Sse a eS e cie korea 314 Nitrogenous bodies (meat proteins), separation, recommendations by Com- mittee Brae nsseccee ance 333 recommendations by Emmett. 278 report by Emmett............ 267 Nominations, committee, report (Davidson)................. at ay StS 288 Miicerstandyrererees LOLs 1 QUA yess hy ata. oye, ear oe eel hea I Oe eee 346 Oils and fats, recommendations by Committee C.........................-.-5- 284 RECOMMendatlonseb ya WELowee Meee ei ete ae eee 186 repontiby [Kem kote ect tc Hee ohm ans EEE eee 181 Palkin and Emery, paper on estimation of antipyrin, reference................ 848 JPEVOET, CRE muh e Se sbom oe Become Eoebas aaaeeSsho as GAcwonmAEEonbaneenoe 169 Patrick, recommendation on dairy produces bf icShch cctainh Bie artes eee Ree, Aiee BENR IO ERROR 289 Phosphorus, in foods, recommendations by Committee B................... 333 report, by; Horbes) and: Wiuissown..-2- seein eeeee 221 Preservatives, recommendations by Committee C..................-...-220-: 287 recommendations) bys Secken ieee er eel heeeencen heer e eee 218 TE POLUD VASE KET yer e eescal-ey- See ee eee rae 210 Publication of proceedings, appointment of committee, resolution............. 169 Rather, report on testing chemical reagents...... ie SR. Je Fie bed CAM 317 Reagents, chemical, testing, recommendations by Committee B............... 334 recommendations by Rather..................... 329 report by. Rather: eee eee ree nee 317 Recommendations of referees, Committee B, report........................05- 331 Committee:@irepontsea eerie ear raees 282 general committee, report............ 335 WetereesianGoficers mlOlS—1O14en 28 sae Soniaceeca iene ae raat anes ee ae 346 eRolutons scommitvee meport (Bartlett) cossseeers eee aac cen eenoo. Saccharine products, recommendations by Committee C...................... 282 Siw ohn report onlmedicated soft drinks...4-- asses ereeeemaeee os sce eee 343 MEOKEL ETE DOL ti OMSPLESELV.AULVES oc ietcrers clare sie tere eve aie nec oestrone 210 and Clayton, paper on lead in baking powders...................... 264 Seil, report on medicinal plants and drugs, reference.......................... 337 Sri bere portvonnmeaiand nish com ia 4 ance ewan a as SRL lpcha sae Lene ann eae 170 MOUmanMks medicated: report Dy_ ote Ohn-p oceania eee teen sa site eniaeiee 343 Spices, recommendation by Committee C......... 2.00... 0c eee e eee eee 284 Standards; cooperation committee, resolution. we... 22... .2.. 0s eet e ee eee eens 169 lepislatirony suthontzinesmesolutlonsse seers enee ee aoe 169 “SU OIE TayES Noyes oF oXeta | Ohi d Dp g0l adlentero cel cee e Annies 4 Maer oaceat baile ho eee s Sina ume als 208 352 INDEX Sugar and molasses, recommendations by Committee B....................... 332 recommendations bys Crossi-n-scee eater eee ae oe eee 317 report byi Cross /ykearmecechc hoc eee eee 314 Synthetieyproducts, report by merync----n oe dee senna ees cies ee aes 337 Lannin: report by, BacOn:5...2. scene aes sees Se ao ei US eee eee 329 Tea and coffee, recommendations by Bartlett....................0.eeeeee ee eee 208 recommendations by Committee C..................2-.-+.---- 287 reportiby, Bartletta-eep- mince coslocnclceeoeinen eens eee 203 ‘in; in’) foods; report by, lreuthardteepeeeeees eee ee eee ee eee eee 254 Treuthardt, supplementary report on heavy metals in foods: tin.............. 254 Vaniolykerand Wanter, precipitanoniolcasein. se n eer eens eesceeeeenee 281 Vegetables, recommendation by Committee C................0- 000 cence cece 286 recommendation bya Viarruder: erases @aceieen-mee ee ieee eee eee 200 report by Magruder ener emcecer oan eee eee eee 199 Vinegar, recommendations by Committee C...............2002ce0see eee eee es 283 Water, in foods, recommendations by Committee B.......................-.-- 332 recommendations byMicGeer---n4-e ei eee soe 218 report, by: MicGeeni nce: ssc cacti aay tie Seieeics ao ocitehs aera 218 White; reportoncereallproductses.nece-n eee ee eee see eee eee ae eee 195 Wichmann, paper on lime as a neutralizer in dairy products, reference. ....... 195 Wiley, address: references sii Ss accacnrre soos ia ek see en ae 288 Wine; recommendations by, \Committee(Cei4.4-- ce ee es eee 283 Winter and Van Slyke, precipitation of casein.................--.-ee-e eee eee 281 Wussow and Forbes, report on inorganic phosphorus in foods................. 22 PROCEEDINGS OF THE THIRTY-FIRST ANNUAL CON- VENTION OF THE ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS, 1914. FIRST DAY. MONDAY—MORNING SESSION. The thirty-first annual convention of the Association of Official Agri- cultural Chemists was called to order by the president, E. F. Ladd, of Fargo, N. D., on the morning of November 16, 1914, at the Raleigh Hotel, Washington, D. C. The following members and visitors were present: MEMBERS AND VISITORS PRESENT. Abbott, J. S., Bureau of Chemistry, Washington, D. C. Adams, A. B., Bureau of Internal Revenue, Washington, D. C. Albrech, M. C., Lutz and Schramm Co., Pittsburgh, Pa. Albright, A. R., Bureau of Chemistry, Washington, D. C. Alsberg, C. L., Bureau of Chemistry, Washington, D. C. Ames, J. W., Agricultural Experiment Station, Wooster, Ohio. Anderson, F. T., Bureau of Chemistry, Washington, D. C. Andrews, J. P., Office of Experiment Stations, Washington, D. C. Appleman, C. O., Agricultural Experiment Station, College Park, Md. Babcock, J. H., Office of Experiment Stations, Washington, D. C. Bailey, C. H., University Farm, St. Paul, Minn. Bailey, E. M., Agricultural Experiment Station, New Haven, Conn. Bailey, H. S., Bureau of Chemistry, Washington, D. C. Bailey, L. H., Bureau of Chemistry, Washington, D. C. Baker, E. L., Patent Cereals Co., Geneva, N. Y. Baker, H. A., 447 West Fourteenth Street, New York, N. Y. Balcom, R. W., Bureau of Chemistry, Washington, D. C. Baldwin, H. B., Department of Health, 927 Broad Street, Newark, N. J. Balls, A. K., Bureau of Chemistry, Washington, D. C. Bartlett, G. M., Jos. Campbell Co., Camden, N. J. Baston, G. H., Bureau of Plant Industry, Washington, D. C. Bates, Carleton, Bureau of Chemistry, Washington, D. C. Baughman, W. F., Bureau of Chemistry, Washington, D. C. Beal, W. H., Office of Experiment Stations, Washington, D. C. Beattie, J. H., Bureau of Soils, Washington, D. C. Bengis, Robert, Bureau of Chemistry, Washington, D. C. Beyer, G. F., Bureau of Internal Revenue, Washington, D. C. 353 354 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 3 Bidwell, G. L., Bureau of Chemistry, Washington, D. C. Bigelow, W. D., National Canners Association, Washington, D. C. Bishop, H. E., State Board of Health, Indianapolis, Ind. Black, O. F., Bureau of Plant Industry, Washington, D. C. Bonney, V. B., Bureau of Chemistry, Washington, D. C. Bornmann, J. H., Bureau of Chemistry, Washington, D. C. Boyett, W. L., State Food Inspector, College Station, Tex. Boyle, Martin, Bureau of Chemistry, Washington, D. C. Boyles, F. M., MeCormick and Co., Baltimore, Md. Brackett, R. N., State Chemist, Clemson College, 8. C. Bradbury, C. M., State Department of Agriculture, Richmond, Va. Breckenridge, J. E., American Agricultural Chemical Co., 2 Rector Street, New York, N. Y. Brewster, J. F., Bureau of Chemistry, Washington, D. C. Briggs, L. J., Bureau of Plant Industry, Washington, D. C. Brinton, C. 8., Food and Drug Inspection Laboratory, U.S. Appraiser’s Stores, Philadelphia, Pa. Brinton, Mrs. C. S., Haddonfield, N. J. Broomell, A. W., Bureau of Chemistry, Washington, D. C. Brown, B. E., Bureau of Soils, Washington, D. C. Browne, C. A., 80 South Street, New York, N. Y. Bryan, T. J., 4100 Fillmore Street, Chicago, Ill. Buehne, Marie C., Bureau of Chemistry, Washington, D. C. Bunzel, H. H., Bureau of Plant Industry, Washington, D. C. Burnet, W. C., Food and Drug Inspection Laboratory, Savannah, Ga. Burnet, Mrs. W. C., Savannah, Ga. Burnett, L. B., Bureau of Chemistry, Washington, D. C. Callaway, Joseph, jr., Bureau of Chemistry, Washington, D. C. Callenbach, J. A., Churchfield Fertilizer Co., Spartenburg, S. C. Carpenter, F. B., Virginia-Carolina Chemical Co., Richmond, Va. Carroll, J. S., German Kali Works, Atlanta, Ga. Catheart, C. S., Agricultural Experiment Station, New Brunswick, N. J. Child, Ernest, 205-211 Third Avenue, New York, N. Y. Clark, A. W., Agricultural Experiment Station, Geneva, N. Y. Clark, 8. C., Office of Experiment Stations, Washington, D. C. Clay, C. L., Bureau of Chemistry, Washington, D. C. Collins, W. D., Bureau of Chemistry, Washington, D. C. Conover, Courtney, Bureau of Chemistry, Washington, D. C. Cook, F. C., Bureau of Chemistry, Washington, D. C. Coopersmith, S., Bureau of Chemistry, Washington, D. C. Crampton, C. A., Institute of Industrial Research, Washington, D. C. Cutler, Carleton, Agricultural Experiment Station, Lafayette, Ind. Dancy, F. B., Munsey Building, Baltimore, Md. Daudt, H. W., Bureau of Chemistry, Washington, D. C. Davidson, J., Bureau of Chemistry, Washington, D. C. Davidson, R. J., Virginia Polytechnic Institute, Blacksburg, Va. Davis, R. D. E., Bureau of Soils, Washington, D. C. Dewey, G. C., National Canners Association, Washington, D. C. Dodge, C. O., Bureau of Chemistry, Washington, D. C. 19165] MEMBERS AND VISITORS 355 Donk, P. J., National Canners Association, Washington, D. C. Doolittle, R. E., Food and Drug Inspection Laboratory, 641 Washington Street, New York, N. Y. Doolittle, Mrs. R. E., Glen Ridge, N. J. Doran, J. M., Bureau of Internal Revenue, Washington, D. C. Doyle, Aida M., Bureau of Chemistry, Washington, D. C. Dubois, W. L., Hershey Chocolate Co., Hershey, Pa. Dunbar, P. B., Bureau of Chemistry, Washington, D. C. Dunlap, F. L., Victor Chemical Works, Chicago, III. Eimer, W. R., Eimer and Amend, 205 Third Avenue, New York, N. Y. Ellett, W. B., Agricultural Experiment Station, Blacksburg, Va. Elliott, F. L., Bureau of Chemistry, Washington, D. C. Emery, W. O., Bureau of Chemistry, Washington, D. C. Ewing, C. O., Bureau of Chemistry, Washington, D. C. Feldbaum, Jacob, Bureau of Chemistry, Washington, D. C. Fenton, E. H., Bureau of Chemistry, Washington, D. C. Ferris, L. W., Bureau of Chemistry, Washington, D. C, Fetzer, L. W., Office of Experiment Stations, Washington, D. C. Fitz, L. A., Agricultural Experiment Station, Manhattan, Kans. Fitzgerald, F. F., National Canners Association, Washington, D. C. Fletcher, C. C., Bureau of Soils, Washington, D. C. Forst, L. B., Food and Drug Inspection Laboratory, Cincinnati, Ohio. Fraps, G. S., Agricultural Experiment Station, College Station, Tex. Frear, Wm., Agricultural Experiment Station, State College, Pa. Frey, R. W., Bureau of Chemistry, Washington, D. C. Fuller, A. V., Bureau of Animal Industry, Washington, D. C. Fuller, H. C., Institute of Industrial Research, Washington, D. C. Furber, F. B., Bureau of Chemistry, Washington, D. C. Gascoyne, W. J., Gascoyne and Co., Inc., 27 South Gay Street, Baltimore, Md. Gibbs, H. D., Bureau of Chemistry, Washington, D. C. Gibson, A. M., Maryland Agricultural College, College Park, Md. Goodrich, C. E., Bureau of Chemistry, Washington, D. C. Gordon, W. O., Bureau of Chemistry, Washington, D. C. Gowen, P. L., Bureau of Chemistry, Washington, D. C. Grab, E. G., Bureau of Chemistry, Washington, D. C. Graham, J. J. T., Bureau of Chemistry, Washington, D. C. Greathouse, Ruth C., Bureau of Chemistry, Washington, D. C. Griffin, E. L., Bureau of Chemistry, Washington, D. C. Grotlisch, V. E., Bureau of Chemistry, Washington, D. C. Hagedorn, C. F., Armour Fertilizer Works, Chicago, Ill. Hanson, H. H., Agricultural Experiment Station, Orono, Me. Harcourt, R., Ontario Agricultural College, Guelph, Canada. Harris, H. L., Pacific Coast Borax Co., 100 William Street, New York, N. Y. Hartmann, B. G., Food and Drug Inspection Laboratory, 1607 Transportation Building, Chicago, Ill. Hartwell, B. L., Agricultural Experiment Station, Kingston, R. I. Haskins, H. D., Agricultural Experiment Station, Amherst, Mass. 356 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. J, No. 3 Hawkins, L. A., Bureau of Plant Industry, Washington, D.C. Haywood, J. K., Bureau of Chemistry, Washington, D. C. Haywood, W. G., Department of Agriculture, Raleigh, N. C. Hoagland, Ralph, Bureau of Animal Industry, Washington, D. C. Holmes, A. D., Office of Experiment Stations, Washington, D. C. Hortvet, Julius, State Dairy and Food Department, St. Paul, Minn. Howard, B. J., Bureau of Chemistry, Washington, D. C. Hubbard, W. S., Bureau of Chemistry, Washington, D. C. Huckle, Clarence, Bureau of Chemistry, Washington, D. C. Hudson, C. 8., Bureau of Chemistry, Washington, D. C. Huston, H. A., 42 Broadway, New York, N. Y. Ingle, M. J., Bureau of Chemistry, Washington, D. C. Ingersoll, E. H., Bureau of Animal Industry, Washington, D. C. Jackson, F. A., Food and Drug Commission, Woonsocket, R. I. Jackson, R. F., Bureau of Standards, Washington, D. C. Jacobs, B. R., General Baking Co., 30 Church Street, New York, N. Y. Jarrell, T. D., Maryland Agricultural College, College Park, Md. Jenkins, L. J., Bureau of Chemistry, Washington, D. C. Johns, C. O., Bureau of Chemistry, Washington, D.C. Johnson, J. M., Bureau of Chemistry, Washington, D.C. Johnson, M. O., Bureau of Animal Industry, Washington, D. C. Jones, C. H., Agricultural Experiment Station, Burlington, Vt. Kebler, L. F., Bureau of Chemistry, Washington, D. C. Keenan, G. L., Bureau of Chemistry, Washington, D. C. Keister, J. T., Bureau of Chemistry, Washington, D. C. Kellogg, J. W., Department of Agriculture, Harrisburg, Pa. Kerr, R. H., Bureau of Animal Industry, Washington, D. C. Kimberly, C. H., 1319 H Street, N. W., Washington, D.C. Klueter, Harry, State Chemist, Madison, Wis. Knight, H. L., Office of Experiment Stations, Washington, D. C. Knight, O. D., Insecticide and Fungicide Board, Department of Agriculture, Wash- ington, D.C. Ladd, E. F., Agricultural Experiment Station, Agricultural College, N. D. Lang, H. L., Office of Experiment Stations, Washington, D.C. Lathrop, E. C., Bureau of Soils, Washington, D. C. Le Clerc, J. A., Bureau of Chemistry, Washington, D. C. Le Fevre, Edwin, Bureau of Chemistry, Washington, D. C. Lepper, H. A., Bureau of Chemistry, Washington, D. C. Lindemuth, J. R., Bureau of Soils, Washington, D. C. Linder, W. V., Bureau of Internal Revenue, Washington, D. C. Lipman, C. B., University of California, Berkeley, Cal. Loomis, H. M., Bureau of Chemistry, Washington, D. C. Lynch, W. D., Bureau of Chemistry, Washington, D. C. Lythgoe, H. C., State Department of Health, Boston, Mass. McDonnell, C. C., Bureau of Chemistry, Washington, D. C. McDonnell, H. B., State Chemist, College Park, Md. 1915] MEMBERS AND VISITORS 357 McGee, W. J., Bureau of Chemistry, Washington, D. C. McIntire, W. H., University of Tennessee, Knoxville, Tenn. McMillin, H. R., Bureau of Animal Industry, Washington, D. C. Magruder, E. W., State Department of Agriculture, Richmond, Va. Mason, G. F., H. J. Heinz Co., Pittsburgh, Pa. Mastin, M. G., Bureau of Chemistry, Washington, D. C. Mathewson, W. E., Food and Drug Inspection Laboratory, 641 Washington Street, New York, N. Y. Maynard, L. A., Ithaca, N. Y. Mell, C. D., Bureau of Chemistry, Washington, D. C. Menge, G. A., Bureau of Animal Industry, Washington, D. C. Merrill, E. C., Bureau of Chemistry, Washington, D. C. Miller, H. M., National Canners Association, Washington, D. C. Mitchell, A. S., Bureau of Chemistry, Washington, D. C. Moore, C. C., Cosmos Club, Washington, D. C. Morey, C. B., Larkin Co., Buffalo, N. Y. Morgan, H. J., Bureau of Chemistry, Washington, D. C. Morgan, W. J., Bureau of Chemistry, Washington, D. C. Mory, A. V. H., Sears, Roebuck and Co., Chicago, III. Murray, A. G., Bureau of Chemistry, Washington, D. C. Murray, B. L., Merck and Co., New York, N. Y. Nealon, E. J., Bureau of Chemistry, Washington, D. C. Nealy, W. A., Jos. Middleby, jr., Inc., 347 Summer Street, Boston, Mass. Nelson, E. K., Bureau of Chemistry, Washington, D. C. Obst, Maud M., Bureau of Chemistry, Washington, D. C. Padgett, W. L., Bureau of Chemistry, Washington, D. C. Palkin, Samuel, Bureau of Chemistry, Washington, D. C. Palmer, H. E., Bureau of Chemistry, Washington, D. C. Parker, C. E., Bureau of Chemistry, Washington, D. C. Parker, E. G., Bureau of Soils, Washington, D. C. Parkinson, Nell A., Bureau of Chemistry, Washington, D. C. Parsons, C. L., Bureau of Mines, Washington, D. C. Patrick, G. H., Bureau of Chemistry, Washington, D. C. Patten, A. J., Agricultural Experiment Station, East Lansing, Mich. Patten, H. E., Bureau of Chemistry, Washington, D. C. Patterson, H. J., Maryland Agricultural College, College Park, Md. Phelps, E. B., U. 8. Publie Health Service, Washington, D.C. Phelps, I. K., Bureau of Chemistry, Washington, D. C. Phelps, Mrs. I. K., Washington, D. C. Pickel, J. M., State Department of Agriculture, Raleigh, N.C. Pingree, M. H., 2343 South Clinton Street, Baltimore, Md. Piper, H. A., Bureau of Chemistry, Washington, D. C. Pope, W. B., Bureau of Chemistry, Washington, D. C. Porch, M. B., H. J. Heinz Co., Pittsburgh, Pa. Powick, W. C., Bureau of Animal Industry, Washington, D. C. Price, R. L., Bureau of Chemistry, Washington, D. C. Price, T. M., Bureau of Animal Industry, Washington, D. C. 358 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 3 Rabak, Frank, Bureau of Plant Industry, Washington, D. C. Randall, W. W., State Department of Health, 16 West Saratoga Street, Baltimore, Md. Rascher, Charles, Bureau of Chemistry, Washington, D. C. Read, E. Alberta, Bureau of Chemistry, Washington, D. C. Reed, E. O., Bureau of Chemistry, Washington, D. C. Reed, J. B., Bureau of Chemistry, Washington, D. C. Reid, F. R., Bureau of Soils, Washington, D. C. Remsburg, C. G., Maryland Agricultural College, College Park, Md. Rerney, J. T., 1527 New Hampshire Avenue, Washington, D. C. Richardson, W. D., Swift and Co., Chicago, Ill. Roark, R. C., Bureau of Chemistry, Washington, D. C. Robb, J. B., State Department of Agriculture, Richmond, Va. Robertson, B. F., Clemson College, S. C. Rodes, W., Agricultural Experiment Station, Lexington, Ky. Rogers, J. S., Bureau of Chemistry, Washington, D. C. Ross, B. B., State Chemist, Auburn, Ala. Ross, S. H., Bureau of Chemistry, Washington, D. C. Ross, W. H., Bureau of Soils, Washington, D. C. Roth, G. B., U. S. Publie Health Service, Washington, D. C. Rudnick, Paul, Armour and Co., Chicago, Ill. Runyan, E. G., Hutchins Building. Washington, D. C. Russell, G. A., Bureau of Plant Industry, Washington, D. C. St. John, B. H., Bureau of Chemistry, Washington, D. C. Sale, J. W., Bureau of Chemistry, Washington, D. C. Savage, Grace O., Bureau of Chemistry, Washington, D. C. Schreiner, Oswald, Bureau of Soils, Washington, D. C. Seeker, A. F., Food and Drug Inspection Laboratory, 641 Washington Street, New York, N.Y. Seidell, Atherton, U. S. Public Health Service, Washington, D. C. Shannon, F. L., State Analyst, Lansing, Mich. Shaw, R. H., Bureau of Animal Industry, Washington, D. C. Sherwood, S. F., Bureau of Chemistry, Washington, D. C. Shorey, E. C., Bureau of Soils, Washington, D. C. Shrader, J. H., Gibbs Preserving Co., Baltimore, Md. Sievers, A. F., Bureau of Plant Industry, Washington, D. C. Silberberg, Berenice H., Bureau of Chemistry, Washington, D. C. Sindall, H. E., Weikel and Smith Spice Co., Philadelphia, Pa. Skinner, J. J., Bureau of Soils, Washington, D. C. Skinner, W. W., Bureau of Chemistry, Washington, D. C. Smalley, F. N., Southern Cotton Oil Co., Savannah, Ga. Smith, C. M., Bureau of Chemistry, Washington, D. C. Smith, H. R., Bureau of Chemistry, Washington, D. C. Snyder, C. F., Bureau of Standards, Washington, D. C. Soule, A. M. G., Bureau of Inspection, Augusta, Me. Spencer, G. C., Bureau of Chemistry, Washington, D. C. Stallings, R. E., State Chemist, Atlanta, Ga. Stephenson, C. H., Bureau of Chemistry, Washington, D. C. Stevenson, A. F., U.S. Public Health Service, Washington, D. C. Stewart, A. G., 313 Southern Building, Washington, D. C. 19165] MEMBERS AND VISITORS 359 Straughn, M. N., Bureau of Chemistry, Washington, D. C. Street, J. P., Agricultural Experiment Station, New Haven, Conn. Sullivan, M. X., Bureau of Soils, Washington, D. C. Taber, W. C., Bureau of Chemistry, Washington, D. C. Taylor, A. E., Bureau of Chemistry, Washington, D. C. Thatcher, A. S., Bureau of Animal Industry, Washington, D. C. Thom, Charles, Bureau of Chemistry, Washington, D. C. Thornton, E. W., Assistant State Food Chemist, Raleigh, N. C. Tice, W. G., State Board of Health, Trenton, N. J. Toll, J. D., 1010 Arch Street, Philadelphia, Pa. Tolman, L. M., Central Food and Drug Inspection District, 1627 Transportation Building, Chicago, Ill. Trescot, T. C., Bureau of Chemistry, Washington, D. C. Treuthardt, E. L. P., Bureau of Chemistry, Washington, D. C. Trowbridge, P. F., Agricultural Experiment Station, Columbia, Mo. Turner, J. D., Agricultural Experiment Station, Lexington, Ky. Turrentine, J. W., Bureau of Soils, Washington, D. C. Valaer, Peter, jr., Bureau of Internal Revenue, Washington, D. C. Van Slyke, L. L., Agricultural Experiment Station, Geneva, N. Y. Veitch, F. P., Bureau of Chemistry, Washington, D. C. Viehoever, Arno, Bureau of Chemistry, Washington, D. C. von Herff, B., German Kali Works, 1901 McCormick Building, Chicago, Ill. Waldraff, P. H., Bureau of Chemistry, Washington, D. C. Walker, C. H., Bureau of Chemistry, Washington, D. C. Walker, L. 8., Agricultural Experiment Station, Amherst, Mass. Wallis, J. B., American Food Journal, Chicago, II]. Walters, E. H., Bureau of Soils, Washington, D. C. Walton, G. P., Bureau of Chemistry, Washington, D. C. Weaver, F’. P., Department of Agricultural Chemistry, State College, Pa. Wessling, Hannah L., Bureau of Chemistry, Washington, D. C. Wheeler, H. J., American Agricultural Chemical Co., Boston, Mass. White, H. J.. Maryland Agricultural College, College Park, Md. White, J. A., jr., Bureau of Chemistry, Washington, D. C. White, W. S., Room 418, City Hall, Cleveland, Ohio. Whitman, H. A., Bureau of Chemistry, Washington, D. C. Wilbert, M. J., U. S. Public Health Service, Washington, D. C. Wilcox, E. L., Bureau of Chemistry, Washington, D. C. Wiley, H. W., Washington, D. C. Wiley, S. W., Wiley and Co., Baltimore, Md. Williams, C. B., Agricultural Experiment Station, Raleigh, N. C. Williams, H. C., Maryland Agricultural College, College Park, Md. Wilson, J. B., Bureau of Chemistry, Washington, D. C. Withers, W. A., Agricultural Experiment Station, Raleigh, N. C. Wolfe, L. A., Bureau of Chemistry, Washington, D. C. Woodward, H. E., Bureau of Chemistry, Washington, D. C. Wright, C. D., Bureau of Chemistry, Washington, D. C. Wurzbacher, A. F., Baltimore, Md. Young, C. O., Bureau of Chemistry, Washington, D. C. REPORT ON PHOSPHORIC ACID. By A. J. Parren (Agricultural Experiment Station, East Lansing, Mich.), Referee, and L. S. WALKER (Agricultural Experiment Station, Amherst, Mass.), Associate Referee. After carefully reviewing the work on basic slag for the past two years, it seemed unwise to continue along the same lines until at least a general survey and summing up of the situation could be made. In 1911 two methods for total phosphoric acid were tried, the official gravimetric method using (a7) method of making solution and a special method which called for the direct precipitation of the phosphoric acid with magnesia mixture in the presence of citrate of ammonia. Three methods for determining the available phosphoric acid were also tested, namely, the molybdate method, which is essentially the same as the official gravimetric method, the citrate of ammonia-magnesia mixture method, and the iron citrate of ammonia-magnesia mixture method. The results obtained by the various collaborators showed no consistent agreement by these methods. For example, seven analysts by the official gravimetric method on Sample 1 obtained results varying from 17.54 to 17.98 per cent, a difference of 0.44 per cent; two analysts obtained re- sults, 18.22 and 18.60 per cent, a difference of 0.38 per cent, while two analysts working in the same laboratory obtained results varying from 18.88 to 19.25 per cent, a difference of 0.37 per cent. The maximum difference of all the analysts was 1.34 per cent. This same condition prevailed with all the other methods on both samples. In 1912, two methods for total phosphoric acid were used, the official gravimetric method using the (a;) method of making solution and the official gravimetric method using the (a;) method of making solution. For available phosphoric acid the methods tested were the molybdate method, the optional volumetric method, and the citrate of ammonia- magnesia mixture method. In the determination of total phosphoric acid on three slags, the results in all cases did not show close agreement. The method using (a7) solution gave the lowest results, the other method gave on the average higher results in cases when sulphuric acid was used as a solvent than when hydrochloric acid and nitric acid were used. In the determination of available phosphoric acid the results on Samples 1 and 2 do not agree as closely as might be desired, while those on Sample 3 agree very closely. 360 1915] PATTEN AND WALKER: PHOSPHORIC ACID 361 In 1913 three modifications of preparing the solution for the official gravimetric method were tested, namely, (a), (a7), and (a7) plus dehydra- tion. In addition, the optional volumetric method was tried. For available phosphoric acid, the methods used the previous year were again tested. The results obtained by the collaborating chemists showed no better agreement than the previous year. On Sample 3, by the (a7) official gravimetric method, nine chemists obtained results varying from 13.96 to 14.47 per cent, a difference of 0.51 per cent; ten chemists obtained results varying from 14.57 to 14.88 per cent, a difference of 0.31 per cent; one chemist obtained 13.05 per cent and another chemist obtained 15.11 per cent, the maximum variation being 2.06 per cent. This example is typical of the results obtained by the other methods. It is the opinion of the referee that the variations in the determination of total phosphoric acid have been due to the fact that the amount of iron being carried down with the yellow precipitate is not constant. It is possible that manganese may act in much the same way as the iron although this point has not been well established. In order to eliminate the contaminating influence of the iron in the gravimetric method, it was suggested that the addition of a small amount of sodium acetate to the solution before precipitating with magnesia mixture might hold the iron in solution. Preliminary work by the referee along this line gave very satisfactory results and this modification of the official method, was, therefore, included in the instructions sent to the collaborators this year. The influence of the iron in the volumetric method may be eliminated by standardizing the solutions against a standard phosphate material of approximately the same composition as the sample to be worked on. Attention was not called to this point in the instructions sent out this year but it so happened that all the collaborators followed practically this procedure. In all future work, attention should be given to this matter. The citric acid solution of basic slag contains from 95 to 100 per cent of all the silica, 80 to 90 per cent of the calcium and 10 to 20 per cent of the iron. Therefore, the same variations may be expected in the results for available phosphoric acid by the molybdate and volumetric methods as were obtained for total phosphoric acid and, furthermore, the effect of the iron may be eliminated by the modifications mentioned. According to M. Popp, the soluble silica does not affect the results except when the amount of soluble iron is small. On the basis of his investigations on this subject Popp proposed the iron citrate method, which is similar to the iron citrate of ammonia-magnesia mixture method tried by this association two years ago except that ferric chlorid is used 1 Chem. Ztg., 1912, 36: 1102. 362 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 3 instead of ferrous chlorid and, in addition, 10 ce. of 0.3 per cent hydrogen peroxid solution are added to oxidize any sulphid that may be present. Attention has been given to another method known as the Lorenz method, that has been tried out very extensively by the German experiment sta- tions. In this method the yellow precipitate, after filtering, is dried in a partial vacuum and weighed. The results obtained in the preliminary work by both the iron citrate and Lorenz methods gave excellent results and were, therefore, included in the work for this year. At the request of the chairman of the Committee on Availability of Phosphoric Acid in Basie Slag, the work has been conducted on the same slags as used by that committee. Four samples of slag marked A, B, C, and D were thoroughly mixed and sent to eight chemists who had pre- viously signified their willingness to coéperate in the work. The sam- ples were numbered as follows: Sample 1 corresponding to Slag C, Sample 2 corresponding to Slag A, Sample 3 corresponding to Slag D, Sample 4 corresponding to Slag B. In addition to the four slags a synthetic solution, Sample 5, representing as closely as possible the citric acid solution of an average basic slag, was sent out. This solution contained in each 100 ce.: 1.5 grams of ammonium phosphate, 5 grams of calcium chlorid, 10 grams of citric acid, 1.12 grams of ferric chlorid, and 8.8 ce. of sodium silicate (10 per cent). On account of the preliminary work done on these methods by the referee it was impossible to send the samples out as early as was hoped and this fact is probably responsible for so few chemists participating in the work. The following instructions were sent to each collaborator: INSTRUCTIONS FOR COLLABORATORS. Samples 1, 2, 3, and 4 are basic slags used by the special committee. Sample 5 is a synthetic solution containing a known amount of phosphoric acid. The methods outlined under available phosphoric acid only, are to be used for Sample 5. A. Determine moisture in Samples 1, 2, 3, and 4 at 100° C. B. Determine total phosphoric acid in Samples 1, 2, 3, and 4 by each of the follow- ing methods: (a) Official gravimetric method, using (a7) method of making solution (Bur. Chem. Bul. 107, Rev., p. 2). Dehydrate an aliquot (20 cc.) of the basic slag solutions by evaporating to dryness on a steam or hot water bath; take up with 5 ec. of hydro- chlorie acid and 25 ec. of hot water; digest to complete solution and filter off silica (SiO.). From this point proceed as directed for determination of total phosphoric acid (Bur. Chem. Bul. 107, Rev., p. 3). Before precipitating with magnesia mixture, add 5 ee. of 5 per cent sodium acetate. (b) Optional volumetric method.—Determine phosphoric acid in an aliquot of solutions (a7) by the optional volumetric method (b) (Bur. Chem. Bul. 107, Rev., p. 4). (ce) Lorenz method (Landw. Versuchs-Stat., 1901, 55: 183).—Dissolve 2 grams of basic slag in 15 ce. of concentrated sulphuric acid and 5 ec. of concentrated nitric 1915] PATTEN AND WALKER: PHOSPHORIC ACID 363 acid, cool and make up to 200 ce. Into a 200 to 250 ec. beaker measure carefully 20 ec. of the solution and add enough nitrie acid (specific gravity 1.20) to bring the volume to 50 ce. Heat the solution, without stirring, over a wire gauze until the first air bubbles appear. Take from the flame and rotate a few times so that the sides of the beaker will not be overheated and add at once, into the middle of the solution, 50 ce. of the sulpho-molybdate reagent. After the precipitate has settled to the bottom, 5 minutes at the longest, stir vigorously for one-half minute with a glass rod. Cover the beaker and allow to stand overnight. Filter through a platinum or porcelain Gooch crucible using a single thickness of ash- and fat-free filter paper in the bottom. If the paper is cut so that it just fits the bottom of the crucible without turning up on the sides, no trouble will be experienced with the precipitate running through. Wash the precipitate four or five times with 2 per cent ammonium nitrate solution and carefully transfer all the yellow precipitate to the crucible. Then wash the precipitate by filling the crucible once full and twice half full with 95 per cent alcohol and allowing it to run dry after each addition. Next wash with ether in the same manner. Place the crucibles containing the precipitates in a fairly large desiccator (without sulphuric acid or calcium chlorid), exhaust to 100 to 200 mm. pressure, and allow toremain one-half hour before weighing. The weight of ammonium-phosphomolybdate multiplied by the factor 0.03295 gives the amount of phosphoric acid (P20s). Nore.—The solution of basic slag prepared by the (a7) method may be used but in this case the aliquot portion must be made up to 50 cc. with the sulphuric-nitric acid mixture described for this method under available phosphoric acid, Preparation of reagents. Sulpho-molybdate solution.—In a 2 liter glass cylinder add 100 grams of pure, dry ammonium sulphate and 1,000 cc. of nitrie acid (specific gravity 1.36) and stir until the sulphate is dissolved. In a 1,000 ce. flask dissolve 300 grams of pure dry ammonium molybdate in hot water, cool to about 20° C., fill to the mark, and pour in a thin stream into the nitric acid-ammonium sulphate solution. Allow to stand 48 hours at room temperature, filter through acid-resistant paper and preserve in a glass-stoppered bottle in a cool, dark place. C. Determine available phosphoric acid as follows: Concentrated solution of citric acid (10 per cent) —Dissolve in water exactly 200 grams of chemically pure crystallized citric acid having its full percentage of water of crystallization. Make up to exactly 2 liters. (If a large number of analyses are to be made, 0.5 gram of salicylic acid should be added to the liter of this solu- tion to prevent decomposition.) Dilute solution of citric acid (2 per cent).—Mix exactly 1 volume of the concen- trated citric acid solution with 4 volumes of water. The resulting solution should have a temperature of about 17.5° C. when used. Making citric solution—Weigh 5 grams of the basic slag, transfer to a one-half liter Wagner flask containing 5 cc. of 95 per cent aleohol. The flask should have a neck width of at least 20 mm. and be marked at least 8 em. below the mouth. Make up to the mark with dilute citric acid solution (2 per cent) of a temperature of 17.5°C. Fit the flask with a rubber stopper and put at once into the rotary apparatus for 30 minutes, making 30 to 40 revolutions per minute. Takeoff and filterimmediately. Analyses of the citric solution —As soon as the filtration is completed, analyze according to the following methods. Dilute 100 ce. of Sample 5 to 500 ec. and, in every case, use same amount of this solution as is used of basic slag solution. (a) Molybdate method (provisionally adopted 1911).—To 50 ce. of the clear filtrate add 100 cc. of molybdate solution made according to the official methods. Put the 364 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. J, No. 3 beaker into a water bath until the temperature reaches 65° C., take out and allow to cool at ordinary temperature. Then filter, and wash the yellow precipitate of phosphomolybdate of ammonia four or five times with 1 per cent nitric acid. Dis- solve in 100 ec. of 2 per cent ammonium hydroxid (cold), nearly neutralize with hydrochloric acid, and add to the solution 15 ce. of magnesia mixture (made according to the official method) drop by drop during continuous stirring. After 15 minutes add 10 to 12 cc. of ammonium hydroxid solution (specific gravity 0.90), then cover the beaker with a glass cover and allow to stand about 2 hours. Filter the double phosphate of ammonia and magnesia through a tared platinum Gooch crucible, wash six times with 2 per cent ammonium hydroxid, dry and proceed as customary for phosphoric acid determinations. Notre: Better results will be obtainedif a smaller aliquot is taken and it is sug- gested that 20 to 25 cc. be used. Before precipitating with magnesia mixture add 5 ec. of 5 per cent sodium acetate solution. (b) Optional volumetric method —Determine phosphoric acid in an aliquot of the clear solutions by the optional volumetric method (Bur. Chem. Bul. 107, Rev., yoo Gey (dla) ))e (ce) Lorenz method.—This method is conducted in exactly the same manner as for total phosphoric acid with one exception. The solution of basic slag (20 cc.) is made up to 50 ce. with the sulphuric-nitric acid mixture, prepared by mixing 30 ce. of sulphuric acid (specific gravity 1.84) with 1 liter of nitric acid (specific gravity 1.20). (d) Iron citrate method (Landw. Versuchs-Stat., 1913, 79-80: 260).—To 50 ce. of the citric acid solution of basic slag add in succession 25 ce. of the iron-citrate solution, 10 cc. of 0.3 per cent hydrogen peroxid, and 25 ec. of magnesia mixture. Put under stirring apparatus from 15 to 30 minutes, filter, ignite and weigh. It is suggested that the solutions be put under the stirring apparatus before the magnesia mixture is added and that the magnesia mixture be added rather slowly. Tron citrate solution.—Place 1,000 grams of citric acid in a porcelain evaporating dish and pour over it a solution of iron chlorid containing 30 grams in 50 cc. of water. Slowly, and carefully add 4,000 ce. of 20 per cent ammonium hydroxid. After all has dissolved pour into 5,000 ce. flask and when cold fill to mark with water. Filter before using. Magnesia mixture —Dissolve 550 grams of magnesium chlorid and 700 grams of ammonium chlorid in a 10 liter flask with about 2,000 cc. of water. After the solu- tion is complete add 1,750 ec. of 20 per cent ammonium hydroxid and fill to mark with water. Filter after several days standing. PRECAUTIONS AND FURTHER INFORMATION. (1) A photograph and detailed drawings of an inexpensive but efficient shaking apparatus were sent out by the referee for 1911. A copy will be forwarded to any- one codperating in this work this year, on request to L. 8. Walker, Amherst, Mass. (2) The rotary apparatus prescribed for shaking the flasks must not be replaced by ordinary shaking or rocker apparatus as the latter differs in construction and effect. The rotary apparatus must turn around its axle 30 to 40 times per minute. Variations within these limits has no marked influence on the results. (3) The half-liter flasks (after the design of Wagner) must have a neck width of at least 20 mm. and are marked at least 8 em. below the mouth. These two points are important since if the neck width is too narrow and the mark too high the result 1915] PATTEN AND WALKER: PHOSPHORIC ACID 365 will be too low, owing to the movement of the liquid being so limited. (The proper flasks are listed in E. & A. Catalogue, (1913) 3172.) (4) The filtration must be done immediately after 30 minutes’ rotation, and it is best to use the folded filter paper of such size that the whole quantity of the liquid can be poured onto the filter at once. Small and bad filtering papers give rise to error in consequence of too slow filtration. If at first the filtrate is not clear, it must be filtered again (through the same filter) until it becomes clear. (5) If the beaker containing the mixture of phosphatic and molybdie solutions is put into the water bath until the temperature reaches between 60° and 70° C., a precipitate free from silicic acid results. If heating is continued for a much longer time the precipitate will often be mixed withsilicic acid, especially when the molyb- dic solution is not added to the filtrate immediately but only after 6 to 12 hours (or longer) after filtration. If silicic acid is present, the precipitate dissolves slowly in ammonium hydroxid, but at first not clearly. Special attention must be paid to the point that the yellow precipitate is dissolved quickly and quite clearly by ammonium hydroxid (2 percent) not made warm. If the solution becomes clear only after some time, molybdic solution and nitric acid must be added to same in order to get a pure precipitate of phosphomolybdate of ammonia, that is, the phosphoric acid must be reprecipitated by the molybdie solution. RESULTS OF COLLABORATION. Comparative work on basic slag. TOTAL PHOSPHORIC ACID AVAILABLE PHOSPHORIC ACID 2 2 2 eS) a 8 3 5 g : | 2 |i : SAMPLE AND ANALYST a 3 3 ‘=I 4 3 e & i} ao 3 ao 5 p aa ad N 2 ad N an & 30 | £3 8 > Se 8 = SAS |S ne Se earl SAMPLE 1: per cent | per cent | per cent per cent per cent | per cent |per cent | per cent C.§S. Lykes, Clemson Col- | 0.16 | 13.27} 13.30) 13.22 | 12.77) 12.25] 12.65) 12.77 NR Cray donee valetoreve archets sy| ores aed ereretg) ae| Custeroe tillers. alana cae 12.73] 12.34! 12.50) 12.70 Ret eG acral abespal Rete merce 12.37) 12.39] 12.43)...... secrets | Bekeretaonl| Sock Soothe oschehercualcts 12R45 23862265 leer PANEL ARO afctatetafalcccteveie = sai 0.16 | 13.27) 13.30] 13.22 | 12.58} 12.34) 12.56) 12.74 O28: cr East Lan- | 0.28 }...:.. SOS erelorsl Oln|treratencr 12.75} 12.80) 12.66 SUT Pe VEL Chis anys srarereis ereteleys ms By In making up these tables from the results as reported, it was found that E. E. Vanatta had given the figures for permanganate-soluble instead of permanganate-insoluble in the column headed Permanganate-insoluble, at least it appeared so, and the referee took the liberty of making the correction. Attention should also be called to the fact that Parkins and Robb did not use the exact amount of material calculated from the water-insoluble organic nitrogen found, but very closely-approximating quantities, and they used the same amounts of each sample for both the Jones and Street methods. COMMENTS OF COLLABORATORS. G. L. Davis: Water determination a single one, while the total and ammonia nitrogen are averages of closely-agreeing duplicates. Methods outlined by refuree were followed in all particulars. J.J. Taylor: Much difficulty was experienced in getting agreeing duplicates by the Street method, especially with Sample 6. All results fairly concordant by Jones method, except with 1, 6 and 8 which, however, gave somewhat lower but con- cordant results when about 2 grams of ground phosphate rock were mixed with the sample before washing out the water-soluble nitrogen, as recommended by Jones 1915] BRACKETT AND HASKINS: Comparative results on samples submitted for codperation. TABLE 1. NITROGEN 385 STREET METHOD & JONES METHOD zZ a a Oo o 5 a ao | 38a |e SAMPLE AND ANALYST = 5 9 a aie B e 5 a = < BR la @] ose 5 a z to | obs | 238 See) Se ese eee 2 | & | 2 | 8 |2e8| eee SAMPLE i: per cent\per cent|per cent|per cent|per cent TARTAR POMACE (Ci. Ib DENG Ge ae eee crt 5 6.40} 2.52) 0.10 | 2.45 | 0.91} 37 GepbepAnderson..:.c-sseeeeen 11.77) 2.70) 0.14} 2.46 | 0.93 | 38 do dja 4h) Soap Se a occ 12.26} 3.07| 0.17 | 2.71 | 0.838 31 RAB SD CCMEr. .\.5.caccce sees 5.30) 2.69) 0.15 | 2.47 | 0.84 34 WimreRodes: 2 o..00.15 acne 14.06} 2.72) 0.10 | 2.64 | 0.79 30 BR MODCY: 1 S.-755 «sis «eee 12.67| 2.76! 0.10 | 2.42 | 0.73 | 30 T.D. Jarrell and C. G. Rems- | CULTS atu MOEA RES EL aacllneaoo. 2.52) 0.13 | 2.40 | 0.68] 28 ripe) eraskins.). o—\2..0.:,-62 ee 14.14) 2.48) 0.10 | 2.388 | 0.90 | 38 OMEE Jensen... 2'5,.400 asa 7.08) 2.54) 0.15 | 2.46 | 0.78 32 Tals SM cO) ib foc ene nes p 9.50} 2.46) 0.15 | 2.41] 0.92] 38 idiy 10s WEES Oe aepeedbees baer 12.13) 2.50) 0.12 | 2.41 | 1.08 45 GARB OMEZ! ooo. sso a ceucnes ae 6.33] 2.57) 0.11 | 2.388 | 0.74] 31 eelieeMiitChell. jes. vi scence 5.66] 2.58] 0.13 | 2.35 | 0.82] 35 Ibe ANG IER (at eee eased ladoaos AWN oo oot DAD INES arveralla tees Uo Jal. Jeb yd Shit paeemonimrciceate soul lorie op P20) ee aon 2.55 | 0.96] 38 eee RODD eisancme en coon eee Zellers ste 2.58 | LOLSOl| se SAMPLE 2: DRIED BLOOD CGPPG DD AVAS Ed oo. ecelcocynenaier 15.60] 12.60] 0.19 | 12.11] 9.21] 76 Gry Andersons ssa. =52c0-0e6 15.90} 12.66) 0.21 | 12.18) 8.81 | 72 (Gly IPS INGE RE es aan a alls oan bam aS DESO lace cllicersec RR RA YLOD: 5a) s:cjecs:= scare isisysialete 14.85] 14.18) 0.22 | 13.56)10.28 | 76 RPS ePDGEMEL jac. asec eee 14.86] 12.82) 0.25 | 12.01) 9.40 | 78 \ivitinadl 007s (es: BIR e eR OICRIIA IO ec 16.88} 12.38) 0.15 | 11.98] 6.78 | 57 PBS e A NODEY:. c:a:3 oxo ccs ereratere 16.13} 12.30) 0.16 | 11.73) 8.63 | 74 T.D. Jarrell and C.G. Rems- DUE Peek cee sie aisle arene ste citar | Merete 12.41) 0.23 | 11.89] 5.33 | 45 IGM asking. Scie qisrrcc see 16.65) 12.52) 0.19 | 12.17) 8.59 | 71 WME VCNSEN s.:--c.00< coe ree 16.32) 12.52) 0.21 | 12.15) 8.86 | 73 ieise Chilton’. .:...<.scesrannee 16.16} 12.16] 0.38 | 11.93) 8.65 | 73 lde EAE Beeeeadorbmocs oc 16.29) 12.38] 0.19 | 12.02) 9.97 | 88 Grebe Boltz.:.. <2 beeen 16.21] 12.57) 0.20 | 12.21) 8.00 | 66 epee Mitchell «2... se cher 16.60) 12.65) 0.21 | 11.76) 9.31 | 79 Nap Ae errd INS =... 2).\.2, cee eee 15.21) 12.65). .-... IE }a7 A ll reecera| enescice® ere arkins: cs: «costes ere POP S3 le cress 12.41] 8.84] 71 din 4343 184) 6) oMp RRR cas coc el loammerne IEC Gooeae 12531||'9232)|) 76 SAMPLE 3: NINE BY WEIGHT TARTAR POMACE TO ONE DRIED BLOOD (Cy UR AD at ee qneeaee encore: oo 6.95} 3.63] 0.11 3.50] 1.60 46 Ga BS Anderson: 35.002 cseeee 12.06) 3.64) 0.14} 3.51) 1.61 46 Genk w Andersons -h hac cee teee ae Se ae slo eRell| ere hale rcteecye eases dis (aE a eemeea pees aac ote 12.70} 4.18| 0.18 | 3.78) 1.54] 41 organic nitrogen Permanganate in- soluble per cent OoOrFrFoooo FRrOrFoOo mtr Oo LO 00 SCOPPONWN HPONOoOn OHH WO MOO Ce One oocnwn or ae oo 2.03 water- Activity insoluble organic nitrogen 386 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 3 TABLE 1—Continued. z JONES METHOD | STREET METHOD Zz . 5 & 20 |oga) oh Sait a SAMPLE AND ANALYST 2 iS g = BS = BS Sic =6 Vom lime | eet oda PPS csc Psneale a | 2 | 8 | ee |2 8 | See ee leer g 2 g £6 |33a) 28a | Sea | esa SAMPLE 3 -—Continued per cent|per cent|per cent|per cent|per cent per cent 13s 1B DYeiaeXe) Pe gaeo oe esoadeods 5.70) 3.85) 018] 3.41) 1.40) 41 Was | ay Wan RO CeStacranierrct eter earners 14721) 3.72) 0.13 || 3257) 1227) 36 1.26} 65 By Ri) Dobe yidjacin casinos 11.31) 3.40) 0.10} 3.32) 1.26] 38 | 2.00; 40 T. D. Jarrelland C. G. Rems- burg. SoC ‘ilsieseties 3.49] 0.14] 3.34) 1.09] 33 | 1.04] 69 185 1D); Sreseace cae oe on cere 14.40) 3.54) 0.13] 3%38} 1.58) 47 | 0.89] 74 OS WS Jensenss-pers-t etree 10.74, 3.64) 0.14] 3.48} 1.72) 40 | 0.96] 72 Hiei Ses Chilton’ peeeeeeeree ae 10.18} 3.48} 0.19} 3.46) 1.54) 45 1.12) 68 Beda Mana tt asc meecer ei: 12.29) 3.57) 0.13 | 3.32] 2.02) 61 TAO) || 52 GAB IBOltZ nes eonc cco sise isis 7.56] 3.63) 0.12 | 3.43) 1.27) 37 | 1.52'| 56 dip 185 Mita ER Sabon saGAnar 6.91} 3.60) 0.13 | 3.32) 1.32) 40 | 0.96) 71 ospAR MEd PINS hs ae scien te tees flees Seal evaresieke leis eee Bros hee ae Merres 0.49 | 86 DP wEls Parkins peewee ricerca shastoke O206|peeeie 3.47| 1.47) 42 LOR GS A fits] B31 8%(0) 0 oS Base Peeees atest ae eat ee es 8 Beet] |Saaece 3.54] 1.26) 386 | 1.19] 66 SAMPLE 4: MIXED FERTILIZER ABOUT 3.30 PER CENT NITROGEN (2 PER CENT FROM NITROGEN MANURE; 1 PER CENT FROM DRIED BLOOD AND 0.30 PER CENT FROM TARTAR POMACE) Gales Davis {cco poetiate 9.25] 3.34) 0.06 | 2.68] 1.68] 63 | 0.23] 91 (GeebewAndersonseenreeeee ee 1O%52), 3243|(0209))) 2281)" 178)! 63) | -.ros ae GSP Anderson tonne sake | perro Geeta eerste AAW Also loners 0.32] 88 JAJe Naylor soasses aan oe 9.25) 3.74) 0.14 3.04, 2.01) 66 0.38 87 Re Be eemeryceter sree ee S45 Se 4s OLS 2282 Ae 76) 162i Ons yelled Wire hvod esae caren cee 10.86) 3.41) 0.10} 3:05) 1.53) 50) | 0252/1) 32 BSR NODC Yast cence ele eee 9.28} 3.32] 0.05 22051) L267) (63: 0.07 97 T. D. Jarrell and C. G. Rems- [SULne ey Peete on memo aes Sn Selle cla es Sab2OI Wy 22521) 10697 385, sommes He SHaskins sea eee 11.82) 3.30] 0.11 2.80! 1.78) 63 0.34 88 OME Jensentean eck neoneo. 9.41) 3.28] 0.18 | 2.59) 1.60) 62 | 0.37) 86 jal, Si, Clomtioiasesaoadsocsnscas GF59| eoe22|Ol13 2.49] 1.46) 59 0.40 84 Be eVaAnahusen sacs onder OROSlimar2 ol) OLO9) 2259 ele SOlGg, 0.76 71 (Gh, INL eohiAsio Sb oobnasnooacee 8.97] 3.44, 0.10} 2.64 1.27) 48 | 0.61] 77 JeeeieeMVintchell ee ena. ee ceeee 9.30} 3.44, 0.10] 2.70) 1.38} 51 | 0.34] 88 je PAG VENI mins ee ee ae ee CEI) B-Bilipegoses DRGAI Ey soon eae 0.34] 87 OL SHe a Barkin sensei vostevs ae catevenilleratecse ss S485 5 oats 2.97; 1.94) 65 OF32 89 JMB eRop ba bear ertaseceeeese leeraetee SAMall, coe of 3.02} 1.67] 55 | 0.46] 84 SAMPLE 5: NITROGENOUS MANURE ABOUT 8 PER CENT, NITROGEN Gb Davis jade eee 8.78] 8.31] 0.18 | 6.52| 4.11! 638 | 0.57] 90 Gab Andersons.n 7a cece O25 Seo OL235| 16 2d0|) 421210 Gone eee errs Gab Anderson. sees toell te reel ocelot BR O8l russ acetic 0.75 | 88 Dee Mayor ne tecs cece eae 9.30) 10.38} 0.26 | 6.63) 4.26] 64 | 1.06} 84 RerB Deemer enoce comer 8.02} 8.30) 0.23 | 6.17) 3.97] 64 1.07] 88 Wim. Rodestee tt Ween sees 9.65| 7.81] 0.18 | 6.60) 3.29} 50 | 1.43} 80 BR Obe yao ne ate erne 8.79] 8.00) 0.16 | 6.28) 3.72) 59 | 0.386) 94 T. D. Jarrell and C. G. Rems- DUDES Rese eee ere ee 8.08} 0.23 | 6.16} 2.41) 39 | 0.76] 87 ii i i, i 1915] BRACKETT AND HASKINS: NITROGEN TABLE 1—Continued. 387 L f°} a | 8 lege 3 ei dl ait SAMPLE AND ANALYST 2 z Se eee) |) RE Pp a z J ° 5 = g Ee, 9 6 5 25 = a = e SAMPLE 5:—Continued per cent|per cent|per cent|per cent Hee) Sas Kin s' sf s.c1- cycioclec ees 10.28} 8.12) 0.22 | 6.70 ORB J ENSEMs 5. a2, ySeeees 9.08} 8.20) 0.26 | 6.30 Hegse Ohilitont a. ..sesecececie: 9.22) 8.08) 0.25 | 6.38 bye be Vanattaces. scene: 8.78} 8.00) 0.20] 6.11 Geb IBolitiz.. 2). eicaeees meee 8.79) 8.21) 0.22 | 6.20 JeoHew Vint chells.. ocean eee 8.28] 8.34) 0.23 | 6.03 Ib e\edl a his Fane ppdearcoood s Sell SeL6|. see. 6.26 epee Parkins. .'s..cscseemes eee Oerdboeous 6.57 Uo 1B) Rios) He aRBnERB on oncodcidallbbosion Ee Vlo de woe 6.50 SAMPLE 6: NITROGENOUS MANURE ABOUT 6 PER CENT NITROGEN (CRB PDAVIS :Gsaiscis one setae 10.48] 6.59] 0.37 | 5.40 (GapbeeAndersomns.s sensei 11.03) 6.63} 0.39 | 4.79 GaphyrAndersont. cacceins eee aoe Seno alten 4.81 Dende avlor:.i.Jca8 scl te eee 10.55) 7.25) 0.42] 5.16 Ree Deemer; js 8 elo] goo] Foo zo SS |/2su r= ee iS} eos |eae| sas| see Bez @ 2Ok| S5k |] 2 oR f< | 509 |] 825 56a | 5 Se BO Zea) enc | Pam!) 258 per cent|per cent per cent BEET ROOT MANURE THREE PARTS BY WEIGHT TO NITROGENOUS MANURE ONE PART GaL.. Davisic.ccsicsesecee sie: 11.83 G-eb. Andersonta- rere 13.47 Gi Andersoning..3 a0 soccer elles eee lances J: Je, Baylor estes ctleen see 11.30 R.vB Deemer pecceccn ee 10.82 Wm Rodesssc2 50s once enter 12.75 BgRSPODey scents pee eos ae 12.40 T.D. Jarrell and C. G. Rems- lhe 28 Sooeteh aboomeenaaecne bodes HD milaskinsse.canee accent 13 .22 OME Uenseny cos.s..sineee ei vaes 11.95 HesaChiltonssteecse. as -cne 11.87 BE VANALUA En. weiss cloerine 12.12 Gab Boltzsccsessccecescenes 11.81 Jee Matchelli inc cases eres 11.64 lbp J\G LEDC aint aceoosnodesne 11.15 JeeHe Parkins? 22. Tae soe ancooe JB RODD ieee oatelsisrnace eave leone ae 6.01] 1.56 6.11) 1.60 6.69) 1.84 6.16) 1.55 6.30) 1.29 5.78) 1.56 5.92) 1.49 6.06) 1.76 6.04) 1.60 5.80} 1.75 5.95} 1.57 6.07} 1.58 5.95} 1.62 G200\2 eter Ga eee 6.18]...... TABLE 2. 4.27| 1.16] 27 | 0.23']| 95 1.20} 30 | 0.49} 89 4.06} 2.06} 50 | 0.61] 85 4.46} 0.94) 21 | 0.83} 81 4.20} 1.13} 27 | 0.21} 95 4.17} 1.08} 25 | 0.50] 88 4.24 1.15) 27 | 0.33 | 92 4.14; 1.10) 26 | 0.38} 90 4.27) 1.65} 38 | 0.40] 91 4.05} 1.45) 35 | 0.59] 85 4.24) 1.02) 24 | 0.75} 82 3.99} 1.00) 25 | 0.37} 91 “Mo oscadlcasos 0.62 | 85 4.29] 1.29} 29 | 0.40] 90 4.20) 1.23) 29 | 0.37) 91 Summary: Samples Nos. 1 to 8 inclusive. JONES METHOD SAMPLE NUMBER Reported Per cent good STREET METHOD ANDO PWNr — or Reported Per cent good 16 81 16 81 16 31 16 75 16 75 16 7. 15 87 16 88 AVErage ss. jase ccc | + sears seers 19165] BRACKETT AND HASKINS: NITROGEN 389 TABLE 3. Sample 9: Nitrate of soda, commercial. TOTAL ANALYST MOISTURE NITROGEN RESULTS RECEIVED AVERAGE (Cry Lgl DEW an eeeanernccnceahonceaa te 1.95 15.44 15.44, 15.42, 15.46 (GeeH Anderson. .jos erie cee tall BBO ime | eters te cevste ease cies coisa) S:aleenie Re Bes COMEr:. «sens risiocle selene (?) 0.25 UGSELAD) |W Cee ico RECS eee dio ley Uitnyd tn Seenorins pomorie cmc oocan 1.40 15.85 15.72, 15.80, 15.58 Wine Rodesssancece een 1.56 15.29 15.28, 15.30 BPR RODE Y 25) tainelnemeniea wine alse cies 1.24 LG: 92 eis etararraseeratete toes rt coos siecaislan 1a, 1D) Jae Shs seeeon soopcbasdecuaclloedsadeds 1.5) :56)": Il tscyehee error ee tee rsiaros wi tleve ORB PU CNSEM npn, foc cctv erl he Arrieta ouall race eee 15.47 15.44, 15.49 Hiysa@biltone +s ocsee see eee 2.06 LOT. || pocsae ate atiee eee BE iWanatita. clas Sue. wetectae 1.43 14.85 14.96, 14.76, 14.83 GMBNBOltzie..3scecmnoee an eeooee 1.05 14.89 14.84, 14.88, 14.92, 14.95 BMH wODertsOn:-c. 50 eee eer eae 2.35 D5 SOM, [fee opcie cusceineie ee ero ronan eee dq lak, Jeti digit): faneaeeeeoneg SacecRoDsloosbaoob 15.58 15.54, 15.89, 15.48, 15.40 for raw materials. Commenting on Sample 9—in my opinion this (the zinc-fer- rous sulphate-soda) method is the best method for nitrates. I further believe that more accurate results could be obtained by weighing a larger quantity, say 5 to 7 grams, of the sample into a 250 cc. flask, making up solution and taking out an ali- quot. This would tend to lessen the manipulative error in weighing out such a small quantity of material. R. B. Deemer: Instructions were followed as requested with the exception of a variation in the usual method of titration in the determination of ammoniacal nitrogen on Samples 7 and8. Inthe method of digestion the amount of acid greatly prolongs the time required for distillation of the ammonia, owing to the fact that the alkali necessary to neutralize this acid fills the flask so as to produce frothing unless distilled very slowly. With the digestion carried on at such a low temperature and for a period of only 2 hours, 20 ce. would be sufficient and not make this delay. The percentage of water-insoluble nitrogen obtained on 1 gram of material is slightly higher than that obtained on 2 grams. The filtrate from the neutral permanganate digestion of Samples 6 showed no excess of the reagent; the effect, if any, is not evi- dent in the results. A careful study of the results obtained by the Street and Jones methods shows practically no advantage of one over the other so far as accuracy of results is con- cerned. The latter is the shorter in point of time, and generally speaking is the easier of manipulation. The most serious objection to this method is the distilla- tion of 95 ec. from such a small volume, which necessitates carrying the contents of the flask practically to dryness, causing caking and a large percentage of breakage. This probably does not oceur with some forms of condensers. From some results obtained it appears that the success of this method depends upon the completion of the distillation of the required volume of the distillate in the time specified. The most serious difficulty presented by the Street method is the time required for the washing of the residue from the permanganate digestion. This washing might be hastened by the selection of a grade of filter paper specially suited to this work, but as the method is not specific in this particular a C. S. & S. No. 588 folded filter was used and may account for the slow filtration. T.D. Jarrell: The distillates from Samples 7 and 8 in the determination of the ammoniacal nitrogen were very cloudy, the cloudiness beginning when about 40 ec. of the distillate had come over, but the sharpness of the titration did not appear 390 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 3 to be interfered with thereby. In every determination by the Street method, 1 and 6 decolorized all the permanganate solution used. O. F. Jensen: When Samples 7 and 8 are washed on a filter and the ammonia de- termined in the washings, the results are about 0.3 per cent lower than the results by distillation with the samples in a flask. Results on 9 by the Ulsch method for comparison with the zinc-ferrous sulphate-soda method: 15.64 and 15.70, as against 15.44 and 15.49; the Ulsch method giving about 0.20 per cent higher results. G. E. Boltz: One of the greatest sources of error by either method seems to be in transferring from the filter paper, in case of the Street method, the water-insoluble residue to the beaker, and in the Jones method, the dried residue to the Kjeldahl flask. Is there any reason why an additional amount of permanganate cannot be used to oxidize the filter paper, so that the filter paper can be transferred with the residue, or an abestos filter be used which could be transferred with the residue? J. H. Mitchell: I believe that a careful analyst can get fairly good results with either method, although occasionally results obtained in two determinations run side by side will fail to check by five to ten points. The Jones method is much shorter. In this method there is some difficulty in transferring sample from filter paper to distillation flask, as it is hard to remove all the residue from the paper. B. F. Robertson: The only advantage of the zinc-ferrous sulphate-soda method over the combination method,! is the rapidity of the determination, when only a few determinations are to be made. Too close attention is required where a great amount of work is carried on at the same time, for this method to be satisfactory; and, too, more experience is required in order to obtain correct results than a method for routine work should demand. L. A. Hudgins: Not satisfied with results on Samples 3, 6, and 7 by the Street method. Results by the Jones method were not regarded as accurate and were not reported. J.H. Parkins: By the Street method with Samples 1, 5, and 6 the permanganate solution was completely decolorized, and 3 partly so. DISCUSSION OF RESULTS. JONES AND STREET METHODS FOR NITROGEN ACTIVITY. Sample 1.—At first glance the most striking thing about Table 1 is the very wide variation in the moisture reported, which, however, seems to bear little or no relation to the total nitrogen figures. The figures for total nitrogen are, with the exception of the Taylor result, quite satis- factory, and the same can be said of the water-insoluble organic nitrogen results, as well as for those on ammonia nitrogen. With the exception of the determinations by Jarrell and Vanatta, the results reported for nitrogen activity by the Jones method are on the whole satisfactory. The same may be said of the results reported by the Street method, if the results by Tobey, Vanatta, and Poltz are excepted. Sample 2.—The moistures here are much more uniform. With the exception of the results by Taylor and Chilton, the total nitrogen results are fairly satisfactory. The same may be said of the ammonia nitrogen 1 Bur. Chem. Bul. 107, Rev., p. 8—Gunning method modified for nitrates, with addition of mercury oxid; or Kjeldahl-Gunning-Arnold modified for nitrates. 1915] BRACKETT AND HASKINS: NITROGEN 391 figures, if those of Chilton are excepted. The water-insoluble organic nitrogen results are a little disappointing. With the exception of the results reported by Rodes, Jarrell, Vanatta, and Boltz, the figures for nitrogen activity by the Jones method are satisfactory; and the same may be said of the Street method, if the results of Tobey, Vanatta and Robb are omitted. Sample 3.—The moisture determinations are again remarkable and bear no definite relation to the total nitrogen. With the exception of the results of Taylor, the figures for total nitrogen may be regarded as fairly satisfactory. The ammonia nitrogen results are quite so, and the water-insoluble organic nitrogen figures are highly satisfactory, with the possible exception of those of Taylor again. About nine of the results by the Jones method, varying from 40 to 46 per cent, show a fair agreement; five below 40 per cent are obviously too low, and one 61 per cent much too high. These nine best, and probably the most correct, of the results show that a mixture of nine parts by weight of tartar pomace and one of dried blood will give a mean nitrogen activity by the Jones method. The results by the Street method are still more variable, five being from 71 to 76 per cent, seven from 65 to 69 per cent, and four scattering— 52, 56, 40, and 86 respectively. Seven of these results show that by the Street method a mean nitrogen activity is given when a mixture like Sample 3 is made. Sample 4.—While the moisture results are more uniform, they are still too variable. The total nitrogen figures are again fairly satisfactory, with the exception of Taylor and Robb. The ammonia nitrogen figures are on the whole satisfactory. The water-insoluble organic nitrogen results are somewhat too variable. About nine of the results by the Jones method show a reasonable agreement, the remainder, below 60 per cent, appear to be too low. About twelve of the results by the Street method show a fair agreement, the rest with one exception appear too low. Sample 5.—The moisture and total nitrogen results are again rather variable and bear no definite relationship to one another. The ammonia nitrogen figures are good, and the water-insoluble organic nitrogen figures on the whole good. By the Jones method six results are from 60 to 65 per cent, six from 50 to 59 per cent, and the remaining three 39, 47, and 73 per cent, respectively. The results by the Street method are much more satisfactory, eight being from 85 to 90 per cent, seven from 80 to 84 per cent, and one, 94 per cent. Sample 6—The moisture and the total nitrogen results leave some- thing to be desired. The ammonia nitrogen results are excellent, while those on water-insoluble organic nitrogen are too variable again. With the exception of the results of Taylor, Rodes, Jarrell, Haskins, Jensen, and Chilton, the remaining nine results by the Jones method may be 392 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. IJ, No. 3 considered to agree fairly well, if Jones’ result is taken as a criterion, which seems justified. Twelve results by the Street method, varying from 79 to 87 per cent, show a reasonable agreement, while the remaining six re- sults, from 70 to 76 per cent appear to be low. Sample 7—The moisture figures are too variable. The total nitro- gen results are, with the exception of those of Hudgins, quite satisfactory, and so on the whole are the ammonia nitrogen results, as well as the water- insoluble organic nitrogen results. The fourteen results reported by the Jones method are highly satisfactory and agree well. Thirteen of the results by the Street method are in good agreement, while the Davis result is a little high and the Rodes result too low. Sample 8.—The moisture figures are better. The total nitrogen figures are, with the exception of those of Taylor and Hudgins, very satisfactory. The ammonia nitrogen results are on the whole good. The water-insoluble organic nitrogen figures are fairly good. About eleven of the results by the Jones method are in fair agreement, while the results reported by Deemer, Chilton, and Vanatta are too high and that by Rodes too low, as compared with the rest. The results by the Street method are more uniform and in better agreement, with the exception of those of Davis and Tobey, which appear too high. ZINC-FERROUS SULPHATE-SODA METHOD. Sample 9.—Though a considerable variation in moisture results is appar- ent, there does not seem to be any reasonable relationship between them and the total nitrogen figures; they do not follow the same curve, and, furthermore, the moisture is so small that these figures may be neglected in the discussion. A discrepancy of over 1 per cent exists between the highest and lowest nitrogen results reported in the table. If the Tobey and the Vanatta and Boltz results are omitted as being obviously too high and too low, respectively, as compared with the remaining ten results, then these ten results fall into two groups of five each: (a) 15.55 to 15.71, average 15.62; (b) 15.29 to 15.47, average 15.38. These averages differ 0.24 per cent, while, if group (a) or (b) be considered alone, the results are quite favorable to the method; when all the results are taken into account, it appears that more work is necessary before recommend- ing the method as official, and that the comments of Robertson are fully justified. CONCLUSIONS. (1) The referee should have given specific directions for determining moisture, as no doubt at least part of the differences reported are due to different conditions, resulting from various forms of baths used. It is 19165] BRACKETT AND HASKINS: NITROGEN 393 rather interesting to note that the greatest variations in the moisture figures occur in the samples containing tartar pomace. (2) Previous experience that both the Jones and the Street methods enable one to differentiate between good and poor organic nitrogenous materials is confirmed. An exception is to be noted in the case of Sam- ple 7, beet root manure. It is interesting to note further that in pot tests made by Jones, he obtained results which agree with the low activity shown by his method for this material. This is a confirmation of the previous work of Jones and Hartwell, which indicates that the Jones method gives results more in accordance with pot tests than the Street method. (3) The testimony of the collaborators is that the Jones method is somewhat shorter, but the work reported certainly shows that in the hands of the inexperienced more uniform results and more closely-agree- ing results are obtained by the Street than by the Jones method. The indications are that more experience is required to carry out the Jones _method accurately, and that, therefore, there is a larger personal equation. (4) The results this year, while somewhat disappointing, indicate the possibility of fixing standards of activity, if so desired. The results on Sample 4, confirm the standard for mixed fertilizers recommended in South Carolina, which standard was based on the examination of 1536 mixed fertilizers. The South Carolina standard recommended is, that if the water-insoluble organic nitrogen amounts to as much as one-third of the total nitrogen, this water-insoluble organic nitrogen must show an activity of 75 per cent by the Street method. KJELDAHL-GUNNING-ARNOLD METHOD. By oversight the referee did not send out a special sample for this work. This method was adopted as official by this association in 1908 but has never been published in the Methods of Analysis, as Bulletin 107, Revised, has not been revised since that time. The nitrogens in Samples 1 to 8, inclusive, were determined by this method. An exami- nation of the results for total nitrogen will show the agreement of results obtained by sixteen different analysts, which with the exception of those on Sample 2 are quite satisfactory. As was to be expected Samples 1 and 2, containing respectively the smallest and the largest amounts of nitrogen, show the least and the ereatest difference and the greatest and least average difference. On the whole the results are very satisfactory and indicate that this method gives very closely-agreeing results in the hands of different chemists. As no sample was sent out specially for this work, we have no comparative results to submit by other methods for determining nitrogen, but the 394 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 3 TABLE 4. Total nitrogens by Kjeldahl-Gunning-Arnold method. Ss G y, Ss AVERAGE STS RESORTED OE aNGE Pe eeen EES 1 15 0.30 0.144 | Omitting results by Taylor. 2 14 0.67 0.280 | Omitting resultsby Taylor and Chilton. 3 14 0.49 0.224 | Omitting results by Taylor. 4 14 0.36 0.223 | Omitting results by Taylor and Robb. 5 14 0.37 0.200 | Omitting results by Taylor and Rodes. 6 15 0.46 0.244 | Omitting results by Taylor. 7 13 0.41 0.222 | Omitting results by Hudgins and Rodes. 8 13 0.40 0.233 | Omitting results by Taylor, Hudgins and Rodes. Totalss.--- 112 3.46 1.780 Averages. .|\5. sc. = 0.308 0.159 results in our own laboratory and those given by Mr. Trescot show that this method is to be preferred to any other for rapid and accurate work, unless further investigation shall show that the substitution of copper sulphate for oxid of mercury gives as good or better results. Last Janu- ary I received a letter from A. J. Patten (Michigan), who offered to let me have the results of some recent work done in his laboratory on the Kjeldahl-Gunning-Arnold method with the use of copper sulphate in lieu of oxid of mercury. These results are given here: TABLE 5. Determination of nitrogen by the Gunning-Copper and Kjeldahl-Gunning- Arnold methods (Patten). GUNNING-COPPER METHOD KJELDAHL-GUNNING-ARNOLD 1} HOURS METHOD 1} HOURS SUBSTANCE Maximum | Minimum Average Maximum | Minimum | Average per cent per cent per cent per cent per cent per cent Gelatin. Aes See 15.14 15212 15.13 15.16 15.04 15.09 . Egg albumin (dried). . ease 12.86 12.75 12.80 12.86 12.80 12.83 Peptone.. a ...| 14.49 14.38 14.45 14.49 14.38 14.43 @aseinve acerca tee 13.87 13.81 13.83 13.92 13.81 13.85 Weathers sancti 14.49 14.48 14.47 14.54 14.38 114.46 Meathereesssurnnce ess ce 4.83 4.72 4.76 4.80 4.75 4.78 Shes Grapes emcte oe 6.48 6.35 6.43 6.51 6.43 6.47 Animal tankage......... 6.32 6.26 6.31 6.35 6.29 6.33 Garbage tankage........ 2.99 2.96 2.98 3.00 2.96 2.99 Beefiseraps§-- seesee ce 9.14 8.97 9.04 9.05 8.97 8.99 Castor bean pomace ee 4.66 4.55 4.59 4.55 4.52 4.53 Cocoa.. OGntA 4.01 3.94 3.97 3.97 3.90 3.92 Cottonseed meal........ U3 7.30 7.31 7.29 7.24 7.26 Cottonseed meal........ 6.48 6.40 6.42 6.37 6.32 6.35 Linseed meal.. < 5.64 5.59 5.61 5.50 5.50 5.50 Silage: skpseckeee ees 1.30 1.26 1.28 1.29 1.26 1.27 WM ouTss* ce. < a ie sep ee 2.43 2.41 2.42 2.44 2.43 2.43 Bran at ccre ei eee 2.60 2.55 2.58 2.58 2.58 12.58 Boneimealiicnjae cee ace 3.06 3.02 3.04 3.11 3.02 3.07 Bib aroscpe crease tomers kote 2.41 Det 2.38 2.37 2.36 2.37 1 Two determinations only. 1915] BRACKET? AND HASKINS: NITROGEN 395 TABLE 6. Comparative study of time of digestion by the Gunning-Copper and Kjeldahl-Gunning- Arnold methods (Patten). 1 HOUR 1} HOURS SAMPLE AND METHOD a g g ks a g | th Bat | | EO aes lice, |g pope} a a 2 23 5 # 2 A a = < A a a < per cent |per cent |per cent per cent |per cent | per cent BONE MBAL: Gunning-Copper........... 6 2.18) 2.16) 2.16) 4 2.20) 2-18!) 2.19 Kjeldahl-Gunning-Arnold. . 6 2.25] 2.19) 2.20) 4 2.20) 2.18) 2.19 DRIED BLOOD: Gunning-Copper........... 6 | 14.11] 14.03} 14.07} 6 | 14.24) 14.07] 14.13 Kjeldahl-Gunning-Arnold. . 6 | 14.32) 14.11] 14.24, 6 | 14.22) 14.19) 14.29 CYANAMID: Gunning-Copper........... 6 | 15.75] 15.37) 15.52} 6 | 16.62) 15.50) 15.53 Kjeldahl-Gunning-Arnold. . 6 | 15.62) 15.46) 15.55) 6 | 15.54) 15.41) 15.49 LINSEED MEAL: Gunning-Copper........... 6 §.53) 5.45) 5.491 5 5.62) 5.50) 5.57 Kjeldahl-Gunning-Arnold. . 6 5.62} 5.53) 5.56] 4 5.62) 5.59) 5.61 2 HOURS 3 HOURS BONE MEAL: Gunning-Copper............. 6 2-20), 2218)" 2219) 16 PPI) PAPA) 2210) Kjeldahl-Gunning-Arnold. . 6 2.23} 2.19) 2.20) 6 2-22) (2,20) 2.20 DRIED BLOOD: Gunning-Copper........... 6 | 14.32) 14.15) 14.22) 6 | 14.40] 14.23) 14.29 Kjeldahl-Gunning-Arnold. . 6 | 14.30} 14.19) 14.25} 6 | 14.34) 14.15] 14.26 CYANAMID: Gunning-Copper........... 6 | 15.67) 15.46) 15.55} 6 | 15.58) 15.50} 15.53 Kjeldahl-Gunning-Arnold. , 5 | 15.62) 15.50) 15.55} 6 | 15.62) 15.50) 15.55 LINSEED MEAL: Gunning-Copper........... 6 5.621) (5.00) 5.55) 6 5.56] 5.64) 5.62 IKXjeldahl-Gunning-Arnold. . 6 OO Or DS|| 0-08] 6 5.62) 5.56} 5.58 An examination of the figures given in these two tables shows that the results are very satisfactory and leave little or nothing to be desired either as to accuracy or time of digestion. B. F. Robertson of my own laboratory (Clemson College, 8. C.) has been doing some work recently and reports as follows: I have used the Kjeldahl-Gunning-Arnold method for total nitrogen since 1908, and regard this method as the most satisfactory that has yet been brought out, the only objection being the use of mercury, which fouls the condensing tubes in the distillation. Recently I have made some determinations using copper sulphate instead of mercury, and find this to be satisfactory, but requiring a little more time for the digestion. On21samples with nitrogen content varying from 2 to 15 per cent, the general average was 0.03 per cent higher than when mercury was used. With only this limited amount of work, I am unable to say whether copper sulphate can be used with as much accuracy under all conditions as mercury. It may not be out of place in this connection to call attention to the work of T. C. Trescot, Chief of the Nitrogen Laboratory, Bureau of Chem- 396 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 3 istry, published in the Journal of Industrial and Engine2ring Chemistry, 1914, volume 5, page 914, ‘(Comparison of the Kjeldahl-Gunning-Arnold Method with the Official Kjeldahl and the Official Gunning Methods of Determining Nitrogen.’’ Mr. Trescot closes his paper as follows: The general conclusion from these results is that the Kjeldahl-Gunning-Arnold method with one and one-half hours’ oxidation, except in the case of cyanamids, which require two and one-half hours, gives more concordant and reliable estima- tions of nitrogen than the official Gunning or the official Kjeldahl methods, both of which require from 3 to 4 hours for oxidation, depending upon material. I understand from his paper that Mr. Trescot used mercury and no copper sulphate. RECOMMENDATIONS. It is reeommended— (1) That the zine-ferrous sulphate-soda method for nitrates be further studied during the coming year, with a view of final adoption as official in 1916, and that it be now adopted as provisional. (2) That the Jones and Street methods for the determination of organic nitrogen activity be further studied during the coming year, with the special purpose in view of improving or modifying the manipulations in the conduct of each process, so as to increase the accuracy of the water- insoluble organic nitrogen determinations, and, in the case of the Jones method, to overcome the difficulties experienced by most analysts in the distillation with alkaline permanganate; and that they be now adopted as official. (8) That the Kjeldahl-Gunning-Arnold method be further studied, although already adopted as official, especially as to the use of copper sulphate in place of mercury. No report was made by the associate referee on the spacial study of the Kjeldahl method. ANALYSIS OF NITROGEN IN LEATHER WASTE. By R. Pumures Rost (Mellon Institute of Industrial Research, Uni- versity of Pittsburgh, Pittsburgh, Pa.).! During the past few months the Mellon Institute of Industrial Re- search has been conducting quite an extensive research on new methods of converting leather waste into fertilizer using the alkaline and neutral permanganate methods for the determination of available organic nitro- 1 Not read at the meeting. 1915] ROSE: ANALYSIS OF NITROGEN IN LEATHER WASTE 397 gen for control of the work. Since some very interesting data in regard to the application of the two methods of analysis have been obtained, it was thought that perhaps this information would be of interest to the referee of the work on nitrogen done in connection with the work of the Association of Official Agricultural Chemists. In carrying out some 400 analyses of treated leather for available or- ganic nitrogen, it was noticed that there occurred quite wide variations in the results obtained by the use of the two methods. The neutral permanganate method in all cases gave constant results, but the alkaline permanganate in every case gave lower results than the neutral and there was also a variation between results obtained at different times using the alkaline method on the same sample, although there was al- ways good agreement between duplicates. In using both methods the directions for the work described in Bulletin 162 of the Bureau of Chemis- try were followed rigidly and every precaution taken to insure accuracy. In the beginning it was thought that this variation was due to oxidation of the ammonia in the distillation with alkaline permanganate; accord- ingly, a series of experiments was carried out to test this point, but in every case negative results were obtained. This result was rather sur- prising when the work of Herschkowitsch (Zts. physikal. Chem., 1908-09, 65: 93) is considered. Since it is a well-established fact that ammonia forms complex com- pounds with some manganese salts the next step was to test this sup- position, and it was found that amounts of ammonia up to 14 per cent of the total content of nitrogen were retained by the residue. The method of obtaining the results, which are given, is as follows: The soluble part of the residue remaining after the distillation of the fertilizer with the alkaline permanganate solution (25 grams of potassium permanganate plus 150 grams of sodium hydroxid per liter) was taken up with 100 ce. of distilled water, a little cane sugar added to destroy any potassium permanganate remaining in the solution, potassium sulphid added to precipitate any soluble manganese compounds and the resulting mixture distilled, the ammonia being collected in standard acid. The results were checked by the neutral permanganate method and by reversing the alkaline permanganate method, that is, by digesting the material with the alkaline permanganate solution on a water bath for one hour, filtering through an asbestos mat in a Gooch crucible, washing the residue with distilled water and analyzing the residue for nitrogen by the Gunning modification of the Kjeldahl method for total nitrogen. The results are shown in the following table. These analyses were made in duplicate, no duplicates bemg accepted that did not agree within 0.2 per cent. All reagents were checked for nitrogen and the results corrected accordingly. Nos. 1 to 3 inclusive 398 | ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 3 tL, eae & TOTAL FERTILIZER ee 5 2 z 4 A ete: Potediate 3 3 % a = RESIDUE NITROGEN Heyes o< es eto ROGER eRe ert aroob Gr ssn wiclndcocsc 10.21 48.51 14.59 63.10 Same checked by neutral potassium per- marnieanaternethodsnec ss eeeeenee Serer ree eee ee || ene ee 63.17 Same checked by reverse alkaline potassium permanganate methods soem pe cleerae | erste aie | oeeiner eee | eee ene 63.21 INGO 252) i) 25 eAtsks dete arad Meroe eee enna cies fe 9.87 89.05 9.03 98.08 Same checked by neutral potassium per- MEM CMRI HNL cags aie nesnde ypooseeballesaoasessllocewsesualloswean cee 98.13 Same checked by reverse alkaline potassium permanpranavesmet hod asec erat ein | ene eet eee ta Seer 98.17 NOt S 5p neu accn saeco ee oneness acne ele 9.53 81.47 4.51 85.98 Same checked by neutral potassium per- manpanatermethod rere cai: eminem croak wl hardiness cess eeege teu ocle saree 86.18 Same checked by reverse alkaline Po assium permanganate method.. poe Sa cia onias| prone follock om ato: 86.23 INOS: Ayia Mea ercr spate are, Sc tera iste daa ise ces 5.34 76.53 6.87 83.40 Same checked by neutral potassium per- Man ganas teymethod a ye e) nse a — — ES Combustion in furnace; carbon dioxid weighed, corrected for carbon dioxid in ash: W. Hz. MelIntire | 1.0340 | 2.7030 | 1.2190 Wet combustion, 1 : 15 orthophosphoric acid, potassium dichromate; carbon dioxid weighed: W. H. MelIntire | 1.0390 bo .7650 | 1.3280 Combustion with sodium peroxid; carbon dioxid gas measured: W. H. Sachs | 0.8835 | 2.6577 | 1.097 1 Used 3 grams of soil with 10 cc. of concentrated sulphuric acid containing chro- mic acid at rate of 3.3 grams per 50 cc. 414 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 3 Inorganic carbon diorid. SAMPLE 1 SAMPLE 2 SAMPLE 3 ANALYST SE teh Stacie sas | |S eRN DN EGeEN ISLS oS bo ee Method A|Method B|Method A|Method B|Method A]Method B (McIntire)| (Marr) |(McIntire)| (Marr) |(McIntire)} (Marr) per cent per cent per cent per cent per cent per cent C. J. Schollenberger, Ohio..... 0.0110 | 0.0055 | 1.7380 | 2.3232 | 0.0286 | 0.0044 0.0120 | 0.0055 | 1.7380 | 2.3276 | 0.0253 | 0.0033 Sejegaactslansiste oS (Sek eee Bl eee L447 eee ORO aa W.H. McIntire and L. G. Wil- | 0.0180 | 0.0264 | 1.2680 | 2.2700 | 0.0210 | 0.0308 lis, Tennessee. 0.0160 | 0.0188 | 0.3640 | 2.3000 | 0.0235 | 0.0297 0.0155 | 0.0242 | 1.2540 | 2.2700 | 0.0245 | 0.0297 H. A. Noyes, Indiana.........} 0.0066 | 0.0022 | 1.9000 | 2.0500 | 0.0044 | 0.0022 0.0099 | 0.0022 | 0.8800 | 2.1200 | 0.0033 | 0.0022 Berar ae ersten 18800" scecice/nc|seoactern saree JEcceban eBeoaboe W8400)" fe occas cones e tee |e ae Se a etn, tale 195009... oho.) ection ore eee O. F. Jensen, Michigan........ 0.012 | 0.0060 | 1.5900 | 2.1800 | 0.028 | 0 0.008 | 0.0050 | 1.6000 | 2.2400 | 0.018 | 0.0260 QE008 9 OlO050)eeiaece ae liner sae 0.018 OX004) WOX0070 ie aoeas| (sare 0.019)" ||; shee CY: Sa (eres erie eed eras inher [So 3.5 0.5 5 E. E. Vanatta, Missouri....... 0.0060 | 0.0633 | 1.408 | 2.2529 | 0.238 | 0.0558 0.0060 | 0.0756 | 1.204 | 2.2529 | 0.087 | 0.0570 BROS SNAG Soe Som MOnass |Akamooed mason ass 0.0674 W. H. Sachs, Illinois..........| 0.0098 | 0.0066 | 0.7278 | 2.0810 | 0.0176 | 0.0072 0.0054 | 0.0039 | 0.7872 | 2.1120 | 0.0149 | 0.0066 oe nae 0.0033) }'(07960) || 220460) 5.2 asec Etech Shen ote toes 08004: |. 2.1600! |)......: . 2: Sree meer tial ie Seowiere 027960) || :2=2260))). <-. 72o0| See Sersey Se tere ease ees, eer 2.2570 )\ 5 2.nce See Brieiteselisistece anions 2.2480) |. < 1.16168 = AsoOs. BORDEAUX. The results obtained on the two Bordeaux mixtures were as follows: Analysis of Bordeaux mixtures. SAMPLE AND ANALYST MOISTURE ea (sxnerno- (nin ac DRY BORDEAUX: per cent per cent per cent per cent We Bebllettand eee EG Var ciniaelie ee. se scree oe eee 13.65 ence lac enaneod ooraasc os 13.65 BS tMotetd fai alall Sy steysaysterela, a donee eee 13.65 SO aA) (es He ASIG A IGls- 6 1 13.80 Bieielsy state es euel's ais oe ieceteere | neeteraeens 13.75 Averages: ead Se ac eee oo Oh ck ocho eee aa (hceic tee dl cic, 13.70 O.B. Winter, Michigan............. 6.94 Dn68)) oi|acasierrcere 13.16 6.98 Not en foie nin 13.26 Ee aR eee 5.67 Suh See 13.21 ANCL ARS rei tae Sie oa cbeecn nici ahaa 6.96 DOA. salll wieereeens 13.21 K. L. Griffin, Washington, D.C...... 7.59 5.83 13.50 13.55 7.56 5.85 13.47 13.57 IAVGTA EC vant tinces hi teee ue fenieie 7.58 5.84 13.49 13.56 1915] ROARK: INSECTICIDES 4A7 Analysis of Bordeaux mixtures.—Concluded. SAMPLE AND ANALYST MOISTURE Sane ON (srnerno- (rmiosct- DRY BORDEAUX: per cent per cent per cent per cent Geo. P. Koch, Minnesota........... 5.02 SUGSMy |everstsyss ates 12.62 4.93 DO OM Uiloge be riatete 12.62 chica crete 5.63 Noo on ono 12.58 Rene Maen s cndllobnbhaysek 12.62 JA GLEN SOR ea ge oh ns Sg OD OO OMTEE 4.98 5.644 | cccassverree 12.61 Ba Re Lobey,) Maines ccaer sce: 5.97 5.84 13.44 13.55 Pie oneal dcr mem ae 13.57 13.97 Bote crest ty aayel| (ud oreeructaiesausl| erect teeeete 13.73 SRidbAnsoed depose oon ipabcoor so 13.63 PAGE AL Gs oicrstoscre s. METHOD Dried at 125°] Ignited SUBSTANCE AND ANALYST SOLUTION No. 1:—Continued DEP OR Bemcert EP CED: per cent OBB RWANGEr 5 c8sisic 5 4% ste neces 41.46 41.42 41.18 37.71 A SOA OBOS 41.42 41.28) || :,..teeentee Ba peleieiete area lsoeieteisie eis AU 14) || \.naweceee Sephetereasts «il Sisatnateorsiaw 41512 o|inaninmee Averaged: sien ieee neh ceeeeeindnees 41.46 41.42 41.17 37.71 IMR. /Millerereeescseacee cee eee 41.60 40.96 41.11 40.10 41.40 402900 | oik res tel Ree eee Be BUS C5 LA 0) Fea ieee (orris gncoc.c Averapetactescencctnccn neon saia 41.50 40.96 41.11 40.10 SD pAverittientuckey ry crt cesctl cies celles | estas ein ele | ese ater 41.17 eSyeyovessocrsh tie | sieaeann oe eae | bere oer 40.84 Ee dass epee satirstovere aise eee oem 40.82 SAV CTA RO RE cis. c6 ar aiatsrovspa, ay ssctAe te retell tserarceeotetecal| @ aakstercrs sits eoereneeeaen 40.94 W.B. Ellett and W. G. Harris, Virginia }.......... 40.82 40:89" ae s(sianacekeyoteney® 40.89 41219) || iS eee arses 40290 Wg iclch cos) eos P8217: 3 Spe BONE AMRA mor Hemi icaoconcoad Gacomcod oc 40.87 41.04: |): cee Fir. © Roark: santas ocracaeas a sicccsestctel cre roue | eteteapate te wack 40.86 40.92 40.27 bate teh tees 40.98 41.06. has Saceneee AA VOL B20 oats aratz arsine eiveroi avers ara vareterersie esererel| reece Stay 40.92 40.99 40.27 SOLUTION NO. 2: EMER -SMIOr Pans sc5 ccsrepatelcstovcvelore cies ere eyaner= 16.14 16.00 16.21 15.70 Spore aiateietees 16.42 15.99 Sete acoverafsiae 16.14 15.90 Sealey te toroeinte 16.32 15.97 sheraysbersreyoreys 16.44 16.64 sc dee esas 16.66 16.69 AAVOTARC stersvcienrceie acelv sic terors ere rareysnretaterere es 16.14 16.33 16.23 16.20 OLBs Winter seseys. ckeesacas oe reccne ae 16.92 16.74 16:86) |pcoeneere Hee races 16.86 16:76) 9). Soe JG CW REO CEES O Hon Gadto tn aeeae 16.92 16.80 VGi8l” || Seer Mi Ri pMiller: ccc. ssiaciisccntssers sosciee es teeters 16.76 16.80 17.41 one ata 16.72 16.75 15.18 Bene cenar aatmeoo nos sauce soc 16.21 E ic lasaus sveweyersil fotecanevseoverore es | lone totenerenronene 15.17 AVOLA RE sesso Rcens See Se es aa reer ier 16.74 16.78 15.99 19165] ROARK: INSECTICIDES 455 Determination of nicotin.—Concluded. SILICOTUNGSTIC ACID KISSLING METHOD LLOYD METHOD) |S eee | «METHOD Dried at 125°} Ignited SUBSTANCE AND ANALYST per cent per cent per cent per cent SOLUTION No. 2—Continued (Shy 1D)S AQ ela irl tomer aie GOS ere GORA Taree Se OnGRan ohana oodellonoad tated 17.62 eee See cos onl bcs eaoe 17.59 SOSA rao oisod is coticcaacdd 17.81 PAV CY RGC'-/-fetcrarcis eisisiesayelausisie.eiere: Sele scsiaiaie 6 | ccs ca ses ays. 85s His, srsleccratever sa | estate eer 17.67 Weber bilettrand We Ge. Harris: so. -jo00-|scocsese«- 16.26 16:14 | oaaaasas eeremenciarsreioce 16.36 Fate ieteectcisters 16.20 PAMELA R Chstcivey-1o)aictereieivicle[sisinsnie e(e(eie isso sicie| sloistoitys = axe 16.27 16.42) ||... scsi lity Gio 1Rko Yad Soe a ieee Reo aR SORE ao ae! Ime e cece 16.87 16.58). \\lscceeaceere seat oenoee 16.65 16:79) |S... 5 see J. NIGIE 3 aegenehed CHa ORee SoD Er Ocoonaane las oas teGec 16.76 GAGS | Rrercreretreine Theoretical = 16.42. first two or three extractions. It may then be thinned with a few drops of water. In this almost crumbly condition of the magma the extractant can be made to reach perfectly every part of the mass and five or six extractions are usually sufficient. H. H. Morgan: In Lloyd’s method for nicotin the indicator used, an aqueous extract of litmus, proved very unsatisfactory, the end point not being sharp. In the silicotungstic acid method (b) a considerable amount of the precipitate was washed through the filter (S. & S. 589 blue ribbon). Possibly this could be avoided by reducing the amount of precipitate, or by using a wash stronger in acid. M. R. Miller: Of the methods used, the one which gave the best satisfaction was the silicotungstic acid method. The Kissling method gives trouble in the final titration using cochineal as indicator as it is found that the end point when nicotin is present in the solution is obscure and difficult to decide upon. In the Lloyd method which was submitted the preparation of the sample for titration is much more cumbersome than that of the method of Bulletin 102, of Bureau of Plant Industry, and in the final titration, there is nothing gained by the use of litmus (azo- litmin in this laboratory) as an indicator, the end point in this case being even more obscure than that. obtained when cochineal is used in the Kissling method. In the silicotungstic acid method, there is less chance for the personal equation to affect the results and consequently the results should be found to be more uniform when given by different workers. The results by the silicotungstic acid check very closely, and almost all of the codperators who tried all the methods expressed their preference for this method. It is very accurate and is preferable in many ways to the other methods. It is recommended that this method be adopted as an Official one, and that the precipitate of nicotin silicotungstate be 456 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. 1, No. 38 ignited rather than dried at 125°. The ignition method is quicker, and in general, more accurate. The referee cannot see any advantage in the Lloyd method for the determination of nicotin as regards quickness, and it certainly is not to be compared in accuracy with the silicotungstate method. It is recom- mended, therefore, that no further work be done on it. It should be said that the directions for the Lloyd method as sent out by the referee aimed to overcome the difficulties encountered in using this method as published in Bureau of Chemistry Bulletin 56, pages 114 to 115, which was tried out by the association in 1898 and again in 1902 (Bur. Chem. Bul. 81, p. 203), and found to be inaccurate. Although the modifications introduced seem to have improved the method, still it is far from being satisfactory and in view of the greater accuracy and adapt- ability of the silicotungstate method the necessity for the Lloyd method is not apparent. RECOMMENDATIONS. It is reeommended— (1) That Method I for total arsenious oxid in Paris green, as described on pages 25 to 26, Bureau of Chemistry Bulletin 107, Revised, be changed as described in Method I, page 445 of this report. (2) That Methods II and III for total arsenious oxid in Paris green (Bur. Chem. Bul. 107, Rev., pp. 26-27) be discarded. (3) That Methods II and III for total arsenic in Paris green, pages 445 to 446 of this report, be adopted as official. (4) That the method for the determination of moisture in Bordeaux mixture, as described on pages 436 to 437 of this report, be adopted as provisional. (5) That the method for the determination of carbon dioxid in Bordeaux mixture, as described on page 437 of this report, be adopted as official. (6) That the electrolytic method for the determination of copper in Bordeaux mixture, as described on page 437 of this report, be adopted as official. (7) That the thiosulphate titration method for the determination of copper in Bordeaux mixture, as described on page 448 of this report, be adopted as an optional official method. (8) That the method for moisture in Bordeaux-Paris green, as described on page 437 of this report, be adopted as provisional. (9) That the method for carbon dioxid in Bordeaux-Paris green, as described on page 437 of this report, be adopted as official. (10) That the method for water-soluble arsenious oxid in Bordeaux- Paris green, as described on page 438 of this report, be adopted as provisional. 1915} ROARK: INSECTICIDES 457 (11) That the distillation method for the determination of total arsenic in Bordeaux-Paris green, as described on page 438 of this report, be adopted as official. (12) That the methods of C. M. Smith and C. C. Hedges for the de- termination of total arsenious oxid in Paris green, as described on page 446 of this report, be made optional official methods for the determination of total arsenious oxid in Bordeaux-Paris green. (13) That the electrolytic method for the determination of copper in Bordeaux-lead-arsenate, as described on page 438 of this report, be made an official method for the determination of copper in Bordeaux-Paris green. (14) That the thiosulphate titration method for the determination of copper in Bordeaux-Paris green, as described on page 450 of this report, be made an optional official method. (15) That the method for moisture in Bordeaux-lead arsenate, as described on page 488 of this report, be adopted as provisional. (16) That the method for the determination of carbon dioxid in Bor- deaux-lead arsenate, as described on page 438 of this report, be adopted as official. (17) That the method for water-soluble arsenic oxid in Bordeaux-lead arsenate, as described on pages 438 to 439 of this report, be adopted as provisional, (18) That the electrolytic method for the determination of copper in Bordeaux-lead arsenate, as described on page 438 of this report, be adopted as official. (19) That the method for the determination of lead oxid in Bordeaux- lead arsenate, described on page 488 of this report, be adopted as official. (20) That procedure (b) of the silicotungstate method for the determi- nation of nicotin, as described on page 439 of this report, be adopted as official. (21) That no further work be done on the Lloyd method for the de- termination of nicotin. (22) That the codperative work on insecticides for next year comprise a study of the following: (a) Methods for the determination of As’** and As’ in the presence of each other and in the presence of one or more of the following: lead, copper, zinc, and calcium. (b) A method other than an electrolytic one for the determination of copper, lead and arsenic in a Bordeaux-lead arsenate mixture. (ec) Methods for the determination of the principal ingredients in zinc- arsenic compounds alone and in combination with Bordeaux mixture. The honorary president, H. W. Wiley, made a short address to the association, touching upon the effects of the war in Europe on science and agriculture in this country, especially on the supply of potash needed for our soils. 458 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. 1, No. 3 REPORT ON WATER. By W. W. SKINNER (Bureau of Chemistry, Washington, D. C.), Referee. The results of the codperative work on the determination of strontium which were reported in 1912 and 1913 were so unsatisfactory, the results being uniformly low, that the referee for last year suggested that a critical study of the method be made. In view of the results obtained it was decided to be inadvisable, therefore, to undertake further coéperative work until some conclusion could be reached as to the cause or causes for the low results. The method which has been used in the coéperative work heretofore for the separation and determination of calcium and strontium is that of Stromeyer-Rose as modified by Hillebrand. Briefly, it is the precipitation of strontium and calcium as oxalates, washing with cold water, igniting and weighing the mixed oxids, converting the oxids into nitrates, digestion of the nitrates with an ether alcohol mixture, filtering off the solvent containing the calcium nitrate, converting the strontium into sulphate and weighing. The calcium being obtained by difference by subtracting the calculated weight of strontium oxid from the weight of the mixed oxids. While the results for strontium which have been reported by this method have been uniformly low the calcium which has been obtained by difference has generally agreed very well with theory, indicating that the loss in strontium is probably due entirely to the solubility of the oxalate. The solubility of calcium oxalate which has been pretty fully determined by Holleman, Kohlrausch, and others and later by Richards, McCaffry and Bisbee, is slight, being of the order of 0.4 mg. of calcium in 200 ee. of cold water, the maximum amount of wash ordinarily used in analysis, and 0.9 mg. in water at 95°C. This solubility is sufficiently small to be of no very great significance in ordinary analytical work. The same, however, is not trueof strontium oxalate which according to Kohlrausch is soluble in water at 18° to the extent of 4.6 mg. per 200 cc. expressed as strontium and to the extent of 6.6 mg. per 200 cc. at ordinary temperature, as reported by Treadwell. In order to understand better the significance of the effect of the solu- bility of strontium oxalate it was thought advisable to check these results under actual working conditions as the method is ordinarily employed and also to test the effect of various wash solutions with the object of reducing the solubility to the lowest possible point. Solutions of calcium and strontium salts were prepared and very carefully standardized. For purposes of comparison it was thought advisable to treat calcium oxalate with the same wash solutions which it was proposed to use for strontium oxalate. In each series the following wash solutions were used: 200 ee. of cold water; 200 ce. of hot water; 200 cc. of water charged with carbon 1915] SKINNER: WATER 459 dioxid; 200 cc. of 1 per cent ammonium oxalate solution; 200 cc. of 1 per cent ammonium hydroxid solution. Since some distilled water contains considerable carbon dioxid it was suggested that it might influence the solubility of the oxalates, hence the reason for using this wash. The maximum solubility of calcium oxalate in any of these solutions was 1.2 mg. calcium when hot water was used, which is probably high since the average of four determinations was only 0.5 mg. per 200 cc. Even this slight solubility was materially reduced, however, when a 1 per cent ammonium oxalate solution was substituted for hot water in the washing of the oxalate precipitate, the maximum solubility being 0.1 mg. per 200 cc. while in two of the four experiments the result was zero. The precipitated strontium oxalate was washed in the same manner using the five solutions named above. The greatest solubility was noted for hot water, the maximum being 10.4 mg. of strontium per 200 cc. of wash water. A maximum of 8.3 mg. was obtained when cold water was used as prescribed in the method which shows that the solubility of stron- tium oxalate is of such an order as to cause very serious discrepancies in the results obtained by the method when used for the separation of relatively small amounts of strontium from large amounts of calcium. The solubility of strontium oxalate in a wash of 1 per cent ammonium oxalate solution was found to be very much less than it is when cold water is used, the maximum being 2.1 mg. with an average of 1.1 mg. The solvent effect of the 1 per cent ammonium hydroxid solution seemed to be quite as great as cold water. These preliminary results indicate that the substitution of a 1 per cent ammonium oxalate solution for washing the mixed oxalates would add very materially to the accuracy of the method by reducing the loss due to the solubility of strontium oxalate as well as exert a favorable influence on the determination of calcium. For the purpose of checking the determination of calcium the method suggests the advisability of determining calcium in the ether alcohol mixture used for the separation of the calcium and strontium nitrates. It has been found, however, that the results by this “direct’’ method are invariably low and this is true even when the results by the “indirect ”’ method give figures for calcium which are in practical agreement with theory. It was thought that possibly the small amount of solvent pre- scribed in the method for digesting the dried nitrates, namely, 3 to 5 cc. of ether alcohol mixture, was not sufficient to remove all of the calcium nitrate. A number of experiments were therefore run using various amounts of ether alcohol mixture from 5 to 40 cc. The loss expressed in terms of calcium varied from 1.6 mg. when 5 cc. of solvent were used to 2.3 mg. when 40 ce. were used, indicating that the quantity of solvent used in digestion was not the factor upon which the error depends. It seems [Vol. 1, No. 3 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS 460 pun juanjos fo sjunown burhuna bursn aingxvu joyooy sayja yyun quaugna.g vafo Wnyuons 247 YUN Buyurwuas unroyoo burmoyy “SUL §°0 0} 20° ‘SUL T°0 0} 0° ‘dul 70 09 0° ““UINTO]VO SB IO1Ia UOTYVUIUIOEJap J {4UAA]OS JO ‘00 OF “WNT SB IOMa UOTYwUIUIAYJap T {4UATOS Jo ‘90 CF “UINTO[BO SB 1OII9 SUOT}BUTUTIOJOp E {JUBATOS JO ‘00 QZ UINIO[VO SB IOII9 UOTYBUIUIIA\ap T {JUAATOS JO “09 CT “Burysom yuanbasqns ATM SUOTIBUIUIA}Op 7 UOTyNIOS BIMoUIUIe 4uad Jed ] JO ‘99 QZ SUOTIVUIULIOJOp F UOTNIOS oyvy/eVxO UNTUOWUIG yuUdo Jad T Jo ‘00 COZ SUOT}VUTUIIOYOp Z I9JEM PI}BUOgIv. JO “09 OOT SUOI}ZVUIUIIOJOP J 10}VM JOY JO °09 (OZ SUOTYBUIULIA}Op 9 I94VM P]Od JO "99 QOZ Surysva 10j poss) “SUOYIPUOD Burysom Lapun ayvypxo wnyuons fo hyvyrgqnjoy SUOIBUIULIOJOP fF UOT4NOS VIuOTUWME YUDD Jad |] JO “99 YOT SUOTIVUIULIOJOP F UWOTNIOS o4¥[exO UMIUOWIUI quad Jed T JO “09 COZ SUOI}BUIULIOJOP F I9YVM POYBUOGIBI JO “99 COT SUOTYBUIULIOJOP F IOVBM JOY JO *09 COZ SUOIVUIUMIOJOP F I9}VVM P]OO JO "09 QOZ Burysea 10} posy ‘suouipuos burylom lapun ayoynxo wnrajno fo fiyyrqnjog 1915] WILLIAMS: AVAILABILITY OF PHOSPHORIC ACID IN BASIC SLAG 461 evident, however, from the results that a single extraction with ether alcohol is not sufficient where a high degree of accuracy is desired but that it is necessary after washing to redissolve in water the dried precipi- tate of strontium nitrate, contaminated with small amounts of calcium nitrate, evaporate to dryness, and extract again with the ether alcohol solution. As a result of this preliminary work it is suggested that the method for the separation and determination of calcium and strontium be modified by substituting a solution of ammonium oxalate for the cold water pre- scribed in the present method for washing the mixed oxalates and that at least two extractions of the mixed nitrates with the ether alcohol solvent be made, instead of one as prescribed in the present method. The matter is referred to the referee for next year for further study. REPORT OF COMMITTEE ON AVAILABILITY OF PHOSPHORIC ACID IN BASIC SLAG. By C. B. WrtutaMs (Agricultural Experiment Station, Raleigh, N. C.), Chairman. Since the last meeting of the association, the committee has held one meeting at which the different phases of the work now in progress were thoroughly discussed. The inauguration of the work, of necessity, has been slow chiefly be- cause of the difficulty in getting the soil, selected for the investigations, thoroughly exhausted of phosphoric acid, and to a less extent in finding the land, selected for the field work, to show a lack of uniformity during the preliminary stages of the work to such an extent that in some cases new fields had to be selected for carrying on the investigations. It is thought best by the committee, in consideration of these facts, to reserve the data which it now has in hand from some of those coéperating in the investigations, to present later with such recommendations as seem justified by the results, when more results and fuller reports have been received. There are codperating in the field experiments seven station workers and in the pot culture investigations eleven workers. The committee has up to this time received four preliminary field reports and seven reports from those codperating in the pot culture experiments. During the coming year, the committee expects to prosecute the work aggressively and hopes to make a final report as soon as all the experi- ments have been completed. William Frear, chairman of the Committee on Food Standards, reported that since there had been appointed a committee to codperate with other 462 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. 1, No. 3 committees on food definitions it seemed practical to drop the old com- mittee on food standards. A motion that this committee be discharged was carried by the association. REPORT OF COMMITTEE TO COOPERATE WITH OTHER COMMITTEES ON FOOD DEFINITIONS. By Witu1aM Frear (Agricultural Experiment Station, State College, Pa.), Chairman. An advisory committee composed of one member from each committee of three has held meetings and assigned to each member of the joint committee subjects for study, as evaporated milk, asafetida, cereal preparations, and iodin solutions. Grade standards have been con- sidered but as yet no action has been taken. Certain difficulties are commonly urged against minimum standards which a well-devised sys- tem of grade standards would remedy; the responsibilities and the monetary values involved are great. All sides of each question must be considered as decisions should not be too theoretical or unfair to manu- facturers where slight changes might lead to greater expenditures than are for the good of the public. The committee is down to careful work and good coéperation and is establishing fundamental facts. No report was made by the chairman of the Committee on Editing Methods of Analysis. REPORT OF COMMITTEE ON THE STUDY OF VEGETABLE PROTEINS. By Tuomas B. Ossporne (Agricultural Experiment Station, New Haven, Conn.), Chairman. When my appointment as chairman of the Special Committee on Study of the Vegetable Proteins was announced to me by the secretary he told me that it was the idea of the association that I should plan the work and have it done by others. I was also to appoint two other members of the committee. As Mr. Alsberg and I had already discussed many of the problems involved in a comprehensive study of the vegetable proteins I asked him to join me on this committee and to arrange for further consideration of the work to be done. As it was manifestly impracticable to undertake any laboratory work on this subject during the present year it was not thought necessary to appoint a third member. It was agreed that work in this difficult field of research should not be begun until some means 1915| OSBORNE: REPORT OF COMMITTEE ON VEGETABLE PROTEINS 463 could be found whereby it could be conducted under conditions that would assure its continuation for a sufficient time to enable those engaged in the work to acquire skill and experience in the methods of isolating and separating the proteins from one another, and in the various methods employed in studying their properties. When such experience has been gained it is believed that a large number of the commercially important seeds should be carefully investigated as of the proteins of many of these we have at present no knowledge whatever. These studies should be further extended to the proteins of other parts of plants which are used as food, and of which, as yet, there is no information available. The importance of such knowledge is becoming every day more evident since the study of the relative nutritive value of the few proteins which are available for comparison is revealing wide differences in their nutritive value. We can no longer be content to estimate the food value of vegetable products by simply determining their content of protein nitrogen since the relative proportion of the amino-acids which these yield determines to a very large extent their value in nutrition. Future progress in the study of many important problems in animal, as well as human, nutrition demands that our knowledge of the individual proteins in the different materials used for food be greatly extended. To do this will require a great deal of patient and laborious work, which can be done only by those trained in this special field. There are very few men in this, or any other, country who have had experience in ex- tracting, purifying and studying vegetable proteins, and it will be neces- sary to train from the beginning most of those who are to take part in such studies. To properly and effectively conduct such work will cost more than most experiment stations will probably be willing, or able, to devote to it. It has seemed to us, therefore, a subject which the Bureau of Chemistry was best able to take up, and a very suitable line of research for it to follow. Mr. Alsberg and I have agreed to coéperate in attempting to organize work along these lines, but are not willing to begin until the right men can be secured to undertake it. This has been found difficult, and as a con- sequence we cannot now report any definite plans that are being actually put in operation. We hope, however, that our efforts will sooner or later meet with success and that before long an extensive study of the vege- table proteins will be begun. In view of the peculiar difficulties presented by an effective study of the vegetable proteins we do not consider that this subject is one that can be dealt with successfully by coéperative work by the association. We recommend that the special committee appointed to report on the study of the vegetable proteins be discontinued, but that all the members of the association give their support to work in this field whenever suitable means are found for properly undertaking it. 464 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. 1, No. 3 C. L. Alsberg gave a short supplementary talk, on the need of the iso- lation and hydrolysis of proteins present in seeds and of laying a foundation for feeding and nutrition work. He stated that since it seemed best for this work to be carried on in the Bureau of Chemistry, the subject has been taken up with the Secretary of Agriculture and a small item under biological investigations inserted in the appropriation. The work is being organized in the Bureau and it is hoped that a number of papers on vegetable proteins may be ready for the next meeting. An invitation was extended by C. B. Lipman to the association to hold the 1915 meeting in San Francisco or Berkeley. After the president read the list of cities and organizations sending invitations, the motion that the association meet in 1915 in San Francisco or general vicinity at a date agreeable to other scientific associations, was carried. On the afternoon of the following day, R. N. Brackett made the resolution which was seconded by E. W. Magruder that the next place of meeting be recon- sidered and made a new motion that the meeting be held in Washington. After discussion by B. B. Ross and H. C. Lythgoe showing how much more work is accomplished when the meetings are held in Washington, the new motion for Washington as the next meeting place was carried. No report was made by the chairman of the Committee of Review of the Analysis of Lime Sulphur Solutions. The meeting adjourned at 4.30 p.m. for the day. SECOND DAY. TUESDAY—MORNING SESSION. REPORT ON FOOD ADULTERATION. By Jutius Horrver (State Dairy and Food Department, St. Paul, Minn.), Referee. The work of the present season has been characterized by the comple- tion or rounding out of several lines of investigation which have been under way during the past two years. An unusual number of the associ- ate referees have brought their work to such a satisfactory state that definite recommendations may be made for final adoption of methods or for further study along well-defined lines. The reports submitted tend in the main to indicate a continued interest in the work and purposes of this association; there have been prompt and fairly complete returns from the collaborators and adequate outlines or full reports from the associate referees in ample time for consideration before the date of this meeting. Details of the results accomplished will be fully presented as the referees’ reports are read, but it may be appropriate to refer briefly to a few salient features of these reports. An important advance has been made in the development of the uranyl acetate and ammonium molybdate polarization methods for determining malic and tartaric acids in fruit juices, jellies, sirups, and other fruit products. In the case of malic acid the associate referee proposes a general method applicable to fruit products and other substances, and also recommends the further study of a modified method for citric acid which involves a new principle discovered by L. W. Andrews. Working on this principle, it is promised that a very simple procedure can be developed for the estimation of citric acid. The referee on flavoring extracts has confined the work of the present year to a comparative study of various methods for determining essential oil in extracts of anise, cassia, cinnamon, clove, nutmeg, peppermint, spearmint and wintergreen. The relative merits of the various methods which have been proposed in recent years are quite adequately shown up on the part of half a dozen experienced collaborators. The chief diffi- culties attending the analysis of such extracts as peppermint, clove and nutmeg, it is believed, have been finally overcome, and the associate referee is able to make definite recommendations to the association, which, 465 466 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. 1, No. 3 if favorably considered, will add several new methods which are now recognized as having established value. An examination of the reports received on heavy metals reveals impor- tant progress toward the perfection of methods for the accurate determi- nation of such heavy metals as lead, arsenic, and tin. The so-called modi- fied Gutzeit Method for arsenic has been carefully tried out under various conditions and three important methods for estimating tin have been subjected to critical study. The work on lead has been carried out by the associate referee on baking powder and baking powder materials. No collaborators have assisted in this work during the past year, but the associate referee reports that he has now perfected the modified pro- cedure of Seeker and Clayton in such form as to be prepared to subject the method to a careful collaborative study with a view to final adoption at the next meeting. It is also definitely proposed that methods for copper, zine, nickel and aluminum be made new subjects for study during the coming year. The work of the associate referee on colors has been devoted chiefly to the study of the separation and identification of the coloring matters naturally occurring in fruits. The outline of the work submitted indicates a large amount of original painstaking investigation, and the report on colors is an excellent illustration of the importance of assigning subjects to specialists in various lines and of the great necessity that one man be allowed ample time in which to develop and complete his investigation of the subject under his charge. When the scheme proposed by the associ- ate referee on colors has been perfected and is available for general labora- tory use, our equipment for routine examinations of foods and other products will be greatly increased. The associate referee on saccharine products has taken up a line of investigation which when completed will serve a good purpose. The work of this season has consisted of a study of nine methods for the detection of artificial invert sugar in honey. Samples were submitted to eight collaborators and results which have been received very materially aid us in arriving at a conclusion respecting the relative merits of the different methods, some of which are virtually modifications of the original Fiehe’s test. Some of these tests will be eliminated as a result of the work of the present year, and it remains to be seen which of the nine methods now under consideration will survive after an additional year’s careful examination. In connection with the work on wine it may be confidently stated that the associate referee has succeeded in bringing details of the method pro- posed by Hartmann and Eoff to such a stage that it will be possible to con- sider favorably its adoption at the present meeting. By means of two im- portant modifications, one of which consists in an important change in 1916] HORTVET: FOOD ADULTERATION 467 the manner of using the indicator during titration, the method is brought into such a form that analysts may place increased reliance upon the results obtained for tartaric acid. The work of the associate referee on preservatives has followed out the recommendations approved by the association at its last meeting, and has been confined to the collection of further data concerning the occur- rence of formic acid or what appears to be this substance in natural products, and to a study of the influence of possible interfering substances upon the Fincke determination and the relative value of qualitative tests. This work has been conducted without the assistance of collaborators, but the referee has gone into the subject to such a thorough extent that the association will be presented with results which may be regarded as amply establishing the limits of formic acid naturally occurring in various kinds of products. It is believed that the work of the collaborators of last year upon the Fincke method justifies recommending this method for provisional adoption. The referee is not prepared to recommend a qualita- tive test which can be considered reliable in the hands of inexperienced analysts, and recommends that the association take up a study of quan- titative methods for estimating saccharin. The work on oils and fats has been devoted to the study of two impor- tant methods for detection of phytosterol in mixed animal and vegetable fats. This is a subject which has not heretofore been under investigation by this association, and it is satisfactory to note not only that it has been deemed advisable to undertake this work, but also that the report of the associate referee reaches conclusions which are of value. Of the two methods which have been subjected to comparison, it may be stated that no choice has developed between the two as respects accuracy, correct conclusions being reached by all collaborators, but that the so-called digitonin method while having advantage of simplicity and convenience, has the disadvantage of requiring an expensive reagent which is not easily obtained. The present Bureau of Animal Industry method, while requiring more time and labor, has been found to be decidedly superior to our present official method. Both methods are recommended for adoption as provisional. The report of the referee on inorganic phosphorus in vegetable and animal tissues is a valuable contribution to the program of this meeting. The work has been carried out entirely in the laboratory of the associate referee, but results are nevertheless ample and conclusive, perhaps in many respects more satisfactory than could possibly be obtained by a number of collaborators working separately. As regards the results obtained on animal substances, conclusions point favorably to the adoption of the magnesia mixture method of Forbes and associates for determination of water-soluble inorganic phosphorus. This portion of the work is in 468 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. 1, No. 3 very satisfactory shape and appears to bring to a conclusion the very careful work which has been carried out during the past three or more years on this subject. While it is recommended that a method be adopted for animal substances, the referee has come to the conclusion that the work on inorganic phosphorus in vegetable substances be dropped. The referee on dairy products has carried out the recommendations of this association in respect to the study of the method, with modifications, which was proposed for final adoption at this meeting. A dozen col- laborators assisted in the work, and results seem to be ample in so far as they reveal the fact that we are confronted with certain difficulties which have not heretofore been fully realized. Of the three samples submitted to the collaborators, two were condensed milk products, sweetened and unsweetened, and it is chiefly in connection with results on these two kinds of products that the trouble has been disclosed. Comparative tests were carried out by the official Roese-Gottlieb method, and the results by this method are equally disappointing and tend to further magnify the importance of a special study of methods for determining fat in evaporated milk. Conversation and correspondence with chemists and manufacturers has increased the realization that we are in urgent need of a thorough investigation on this subject. Results reported at the meet- ing of 1913 on samples of milk, cream and ice cream were fairly satis- factory and pointed favorably toward a conclusion of the work during the past season. Assuming that reliable uniform samples were submitted to the collaborators, and we have every reason to rely on that assumption, it is impossible to conclude otherwise than that there are peculiar diffi- culties attending the accurate determination of fat in evaporated and condensed milk. Certain processes incident to manufacture, added to which is the recent introduction of the homogenizer by a large number of manufacturers, doubtless cause important changes in the constitution of these products to such an extent that there are unusual difficulties in separating the fat from the other ingredients by any of the methods with which we are now acquainted. It is, therefore, recommended that this subject be given special study during the coming year. Satisfactory as the work of the past season has turned out to be, we have been nevertheless hampered by an unfortunate state of affairs respecting the publication of our proceedings. This condition has existed or has been in existence more or less seriously for a number of years. You will recall that the published proceedings of the 1912 meeting were not in evidence until the opening day of our session a year ago, and as I am writing this there is much doubt whether we are to see anything at all of the proceedings of the meeting of 1913. Respecting this last supposition, however, it matters little in so far as it affects the real situation; whether we are to be favored with a copy of our last year’s proceedings or not, the 1915] HORTVET: FOOD ADULTERATION 469 general unsatisfactory conditions still confront us—we are not kept adequately in touch with our work from year to year, there is a conse- quent lag in our interest, and this unbusinesslike state of affairs can but reflect unfavorably on this association as a progressive organization. It is safe to venture the prediction that the members of this association would be well pleased with at least one publication annually, and in the main, all would be well satisfied if such a publication of proceedings could be depended on to make its appearance within six months after the con- vention. The work of this association is important; all chemists who have in the past twenty years been in close touch with the State and Federal work in its various branches need no assurances on that point; in fact, all will insist that our work is essential, not only in the interest of the public welfare, but also in the interest of ourselves professionally. No association has worked more disinterestedly for the promotion of things purely scientific or with greater zeal and self-sacrifice for the cause of pure products and high standards of quality. We are, as a matter of fact, vitally a part of the official work of every State in the Union, and con- stitute an essential factor in more than one department of the Federal service. Every State chemical laboratory, whatever its departmental connection, has a vital interest in the development from year to year of the studies being carried on and the results reported by this association. With these facts in mind, may we not readily see a solution of the problem before us? It may or may not be a legal obligation on the part of any of the States or of any Federal bureau or department to make good the deficiencies which hamper our work, but all interests involved can doubtless be assembled in such shape as to formulate a plan whereby we can all rest assured of the continuance of our association as an official or at least in a serious sense, a quasi-official organization, and thereby provide also the necessary means to insure the regular and prompt ap- pearance of our printed proceedings. It is the urgent wish of your referee that this subject be taken up at this meeting with a view to an early practical arrangement whereby we may rest assured in the future regard- ing the proper care of such publications as are deemed essential to our work. Furthermore, year by year our work has before it a definite purpose; we naturally like to see our results from time to time brought into definite form, stripped of nonessentials, and brought down to date. We have on our hands a mass of old matter that it would seem pleasant to discard, methods that are obsolete, data that should be revised or entirely thrown out. Bulletin 107, Revised, is now old; there are six years since its publi- cation, probably seven or eight years since its compilation; and this con- stitutes a long period of time as we are in the habit of measuring time in terms of chemical events. All of us, whether beginners, independent 470 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. 1, No. 3 investigators, or more or less confirmed followers of routine, habitually refer to the Official and Provisional Methods of Analysis; in a sense it is not only our duty as official chemists but it is also often a legal necessity. We have thumbed the leaves of Bulletin 107 so often that its pages have begun to stale, yet we hold no sacred views respecting a document that is so old as to be in many features out of date or at worst a relic of earlier ideals and habits of thought. We are now in a mood to demand a revision of 107, Revised—better, in fact, to require a complete new compilation which shall embody not only all that is good and permanent in the publi- cation now in print, but also all that has been well tried out and finally adopted by the association during the past six years. It is hoped that such revision or rather compilation is now well under way; and that a new bulletin of official methods of analysis will be in our hands at an early date. Therefore, as a second recommendation, I would urge that this association consider the question as to progress now being made on a compilation to date of our official provisional methods of analysis to super- sede Bulletin 107, Revised, and that we arrange at this meeting a definite plan whereby the work may be carried out to completion at an early date, and whereby the expense of publication may be satisfactorily met. REPORT ON COLORS. By W. E. Maruewson (Bureau of Chemistry Food and Drug Inspec- tion Laboratory, New York, N. Y.), Associate Referee. The codperative work on colors for the past three or four years has been confined chiefly to the coal tar dyes. These being of well understood chemical nature the analysts have had at their disposal much reliable information concerning the properties of the individual substances and one of the main problems has been the selection of the characteristics best suited for practical analytical work. By the use of immiscible solvents, supplemented in a few cases by chemical methods, it would seem that most dye mixtures likely to be met may be separated satis- factorily and their identification completed by well-investigated reactions. This year it was hoped that similar methods might be extended to the commonly-occurring natural coloring matters. These present more difficulty because of the limited solubility in organic solvents of many of them, their frequent occurrence in admixture with one another and especially because of the lack of much exact knowledge concerning them. After some experimental work by the associate referee, however, it seemed that little would be gained by sending out samples for analytical investi- gation by methods now in use whose limitations are already known, and a circular letter was sent to the collaborators stating this and expressing 1915] MATHEWSON: COLORS 471 the hope that any new data might be published or be communicated to the other analysts through the associate referee as soon as possible. The experimental work just mentioned failed to bring out distinctions sufficiently marked to be of use in analysis between the coloring matters of a number of common fruits, authentic samples of the juices of cherries, blackberries, raspberries, currants, grapes and huckleberries having been kindly furnished for this work by H. C. Gore of the Bureau of Chemistry. These coloring matters, though extracted in relatively small amount by amyl] alcohol from acid solutions, were found to be rather readily changed by heating with dilute hydrochloric acid, substances similar in color re- actions but much more soluble in amyl alcohol being formed. These derived colors were obtained fairly free from other substances by treat- ment of the fruit juice with excess of neutral lead acetate solution (practi- cally all coloring matter was thrown down), washing the precipitate with water several times by centrifuge until sugar, etc., were removed, solution of the coloring matter by treatment of the precipitate with 10 per cent hydrochloric acid, and extraction of this solution with amyl alcohol to remove substances directly taken up by this solvent. The hydrochloric acid solution was then boiled a few minutes and finally again shaken out with amyl alcohol. The new coloring matter formed by hydrolysis of the glucosid was then extracted in large pro- portion and the amyl alcohol solution was freed from hydrochloric acid, lead chlorid, etc., by washing four or five times with water. This plan of separation offered the advantage that it could be incor- porated with work already done, giving a qualitative method for separat- ing these colors from all those not changed in solubility by warming with acids (especially the coal tar dyes). The coloring matters obtained, how- ever, were very similar or identical in their behavior toward reagents tried. The red amyl alcohol solutions when shaken with caustic soda solution in absence of air gave up the color to the aqueous layer in all cases forming deep green solutions almost instantly becoming brown in the presence of oxygen. The spectra of the red solutions showed a diffuse absorption band in the greenish yellow, those of the green solutions a much more sharply-defined band in the orange red. Recently R. Willstaetter! has carefully investigated the coloring mat- ters of a number of common fruits and flavors. In general he has found them to consist of glucosids (anthocyans) that on warming with acids break down into new coloring matters (anthocyanidins) and sugar, ete. The anthocyanidins form blue phenolic salts with alkalies, red oxonium salts with acids. The facts brought out in this research will be of great value in the development of analytical methods. 1 Sitzungberichte der KGniglich Preussichen Academie der Wissenschaften, 1914, p. 402. 472 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. 1, No. 3 Bearing on the analysis of coal tar dye mixtures, work by Kehrmann, Havas and Grandmougin! has indicated that a number of common basic dyes containing completely alkylated amino groups (for example, methyl violet, methylene blue) suffer rearrangement from the ortho- quinoid to the para-quinoid form on treatment with alkali, this being accompanied by a splitting off of an alkyl group. When such dyes, therefore, are separated from mixtures by shaking out their alkaline solutions with ether, etc., the original coloring matter may not be ob- tained on treatment with acid, but a lower alkylated derivative. Gowing-Scopes? has recently studied the solubilities of a large number of coal tar dyes in the newer chlorhydrocarbons that have become of technical importance. Tetrachlor- and pentachlorethane and di-, tri-, and tetrachlorethylene were investigated in their relation to 45 dyes (as well as a large number of other substances) and separations based on extraction of the dry coloring matters are indicated. The dyes are de- scribed by trade names which in some cases are rather indefinite. It is recommended that the investigation of methods for the separation and identification of the important natural coloring matters be continued by the association. REPORT ON SACCHARINE PRODUCTS. By F. L. Suannon (State Analyst, Lansing, Mich.), Associate Referee. Detection of Artificial Invert Sugar in Honey. The work during the last year consisted of a study of the methods for the detection of artificial invert sugar in honey. A review of the literature revealed the fact that numerous tests have been proposed for this purpose, some of which are modifications of old tests while others are distinct tests in themselves. They are all color reactions and depend upon the presence of furfural or furfural derivatives. In the past, various workers have commented on the reliability of various tests. It seems that there are more or less confusion and contradictory statements in regard to the differ- ent tests, some claiming that one test is perfectly reliable for the detection of artificial invert sugar in honey, while others claim that the same test is unreliable, the consensus of opinion, however, being that there is no single test that can be depended upon unless supplemented by further work. It was, therefore, attempted in this work to select only those tests which had met with the greatest success in the hands of others. Nine of the most common tests were selected and instructions sent out to the collaborators along with the samples. 1 Ber. d. chem. Ges., 1913, 46: 2131, 2802; 1914, 47: 1881. 2 Analyst, 1914, 39: 4. 1915] SHANNON: SACCHARINE PRODUCTS 473 PREPARATION OF SAMPLES. Six samples were prepared as follows: Sample A: Pure honey. Sample B: Pure honey + 5 per cent commercial invert sugar. Sample C: Pure honey + 20 per cent commercial invert sugar. Sample D: Pure honey + 50 per cent commercial invert sugar. Sample E: Pure honey + 5 per cent Herzfeld sirup. Sample F: Pure honey + 20 per cent Herzfeld sirup. The commercial invert sugar was obtained on the open market. The Herzfeld sirup was prepared in this laboratory followmg the method of Herzfeld! “One kilogram of refined sugar is heated with 300 ce. water and 1.1 gram of tartaric acid to boiling and maintained at this temperature until the solution acquires a golden yellow color (one-half to three-quarters of an hour). COLLABORATORS. The following seven collaborators who took up this work, sent in reports: Julius Hortvet (Guy A. Parkin, analyst), St. Paul, Minn.; C. E. Warri- ner, Fort William, Ont.; Miss N. A. Childs, Lansing, Mich.; W. C. Geagley, Lansing, Mich.; W. S. Hubbard, Ann Arbor, Mich.; N. B. Lawrence, Ann Arbor, Mich.; F. L. Shannon, Lansing, Mich. INSTRUCTIONS TO COLLABORATORS. To each sample submitted apply the following tests: 1. FIEHE’S ORIGINAL TEST (Analyst 1908, 33: 397). Reagent.—Dissolve 1 gram of resorcinol in 100 ce. of hydrochloric acid (1.19). Redistilled ether. Manipulation.—One gram of honey is rubbed down in amortar withether. Filter off ether and evaporate. Moisten the residue with one drop of the reagent. Results.—In the presence of artificial invert sugar an orange red color is developed, changing to cherry red and then to brown red. Pure honeys sometimes give a pink coloration. Note color after standing 10 minutes and again after standing 24 hours. 2. REINHARDT’S MODIFICATION OF FIEHE’S TEST (Analyst, 1910, 35: 434). Reagent.—Use 25 per cent hydrochloric acid instead of 38 per cent as in 1. Manipulation.—Same as in 1. Results.—Same as in 1. 3. HALPHEN’S MODIFICATION OF FIEHE’s TEST (Chem. Abst. 1911, 5: 2124). Reagent.—0.02 gram of resorcinol dissolved in a mixture of 2 cc. of dehyrated ether, 25 cc. of absolute alcohol and 3 cc. of hydrochloric acid. Manipulation.—Same as in 1. Results ——Halphen says that his reagent does not give a carmine color with as many flavors as the orignal Fiehe reagent. 1 Bur. Chem. Bul. 110, p. 64. 474 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. 1, No. 8 4. HARTMAN’S MODIFICATION OF FIEHE’S TEST (Chem. Abst., 1912, 6: 1787). Reagent.—Same as in 1. Manipulation.—Add 2 drops of the reagent directly in 1 gram of the honey in a porcelain dish. Results.—If artificial invert sugar is present a cherry red color appears the same as in the original test. Natural honeys give the reaction after standing about 45 minutes. 5. BRYAN’S MODIFICATION OF FIEHE’S TEST (Bur. Chem. Bul. 154, p. 15)- Reagent.—Same as in 1. Manipulation.—Place 10 ce. of a 50 per cent honey solution in a test tube and run in 5 ce. of ether on top. Shake contents vigorously and allow to stand for some time until ether layer is perfectly clear; transfer 2 cc. of this clear ether solution to a small test tube, and add a large-sized drop of the resorcin solution. Shake and note the color immediately. Results.—In the presence of artificial invert sugar the drop of acid in the bottom assumes immediately an orange-red color, turning to a dark red. 6. BROWN’S ANILIN ACETATE TEST (Bur. Chem. Bul. 110, p. 68). Reagent.—(freshly prepared). To 5 ce. of ¢.p. anilin add 5 ce. of water. Addsuffi- cient glacial acetic acid (2 cc.) to clear the emulsion. Manipulation.—To 5 ce. of a concentrated solution of honey (1:1) in a test tube add 1 to 2 cc. of the anilin reagent. Allow the latter to flow down the walls of the test tube to form a layer. Results.—In the presence of artificial invert sugar a red ring forms at the junction of the two liquids. 7. FEDER’S ANILIN CHLORID TEST (Analyst, 1911, 36: 586). Reagent.—(freshly prepared). To 100 ce. of c.p. anilin add 30 ce. of 25 per cent hydrochloric acid. Manipulation.—5 grams of the honey are mixed directly in a porcelain dish with 2.5 ec. of the anilin reagent. Results.—Bright red color indicates artificial invert sugar present. The inten- sity of the color is proportional to the amount present. 8. BENZIDIN ACETATE TEST (Analyst, 1913, 38: 20). Reagent.—Dissolve pure benzidin in dilute acetic acid to a saturated solution. Manipulation.—2 grams of the honey are dissolved in 10 cc. of water and 1 ce. of the filtered reagent. Results.—Artificial honey gives an intense yellow coloration. The depth of color is proportional to the amount present. 9. B-NAPHTHOL TEST, LITTERSHIED (Analyst, 1913, 38: 217). Reagent.—88 to 90 per cent sulphuric acid. Redistilled ether. Manipulation.—Stir 20 grams of honey with 10 ec. of ether, decant and repeat extraction. Transfer ethereal extracts to porcelain dish containing a crystal of B-Naphthol. Treat residue with 5 cc. of the acid. Add acid carefully so that it flows over whole surface of residue. Results —If honey contains artificial invert sugar a Bordeaux red or blue violet coloration will appear within 30 minutes. Natural honey gives a dirty yellow color. 1915] SHANNON: SACCHARINE PRODUCTS 475 Give your conclusions as to the purity of these samples, from the various tests you make on them. Give your opinion as to which ones are adulterated, if any, and which ones are pure, if any. Before submitting the samples and the tests to the various collaborators, the tests were carried out by the associate referee on five samples of pure honey obtained from authentic sources. In the following instead of re- porting the sample as negative or positive, I have indicated the color change that took place. DESCRIPTION OF SAMPLES. Sample 1.—Light colored honey gathered in 1913 from clover: extracted with a power extractor in the cold; flavor good. Sample 2.—Light colored honey, slightly granular, gathered in 1913 from clover; extracted with a power extractor in the cold; flavor good. Sample 3.—Light colored honey, gathered in 1913 from clover; extracted with power extractor in the cold; flavor good. Sample 4.—Dark colored honey, gathered in 1911 from buckwheat and heartsease; flavor poor, similar to sorghum. This sample it will be noted has a tendency to give all the reactions for invert sugar. No doubt due to its age, traces of furfural or furfural derivatives have formed. Sample 5.—Dark colored honey, gathered in 1913 from golden rod; extracted in cold; flavor fair. Results on pure honey. 3 REINHARDT’S MODIFI- ° CATION HALPHEN'S , Ic} ORIGINAL FIEHE TEST (25 PER CENT MODIFICATION ee eS e § HYDROCHLORIC ACID (ETHER SOLUTION) MODIFIGLTION & |(10 minutes to 24 hours)/|(10 minutes to 24 hours)}(10 minutes to 24 hours)|(10 minutes to 24 hours) lias No Faint No Faint No No Faint | Cherry change |_ pink change pink | change } change pink red 2 Do Do Do Do Do Do Do Do 3 Do Do Do Do Do Faint Do Do pink 4....| Faint Pink Faint Pink Do No No Do pink pink change | change 5 No Faint No Faint Do Faint Do Do change pink change pink pink SAMPLE BRYAN'S BROWN'S ANILIN | FEDER'S ANILIN BENZIDIN B-NAPHTHOL NO. MODIFICATION ACETATE TEST CHLORID TEST ACETATE TEST TEST Te Meats ait No No Faint No Dirty change change pink change yellow ae es at Do Do Do Do Do Ghocegere Do Do Do Do Do ee ee tee falll votesete scichareertast Faint Faint Faint Faint red red yellow violet on edges Deane sue No No Faint No Dirty change change pink change yellow [Vol. 1, No. 3 476 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS COCO rer ee OATISOg ai OA1}BION Inyqnog OATIEAON @AIPIBOg od 8A1}BI9 NY isan 10HL -HdVN-& OATHBOT °d (artes aed) BAIPISOg] OATZISOT [Ny9qQn0q| eAtzwFeNy od od od od od od OATPISOg od od PATHISO od od BAIPUAIN | OATIVBIAN od BATPISOT od Tny3qnod [nygqnoqd] 9A1yedaN OATHISOg| 9ATZISOG od od od at od od od od od BATIUION (uMoIq Ystuaadg ) od OATVEION BAIVESAN | OATPVAANT Isa Isat GALVLaovV | ano 1no NIG NITINV “IZNGG s,uadda OAIqGOg (yard) OAIPIBOT (urd) Inyaqnog od od od @AT}ISOg (pea FUIBS) TRANG (rar SATIZBTONT od Injyqnod °d od od BAIVUB NT IsaL TLVLAOV NIUNV 8, NMOUR OATHISOg od O@AIZBION °d og od od OAIPISOT od OAI}VOA9N (urd) [nyaqnod SATPISOT OAIVUBIN BAIZISOg OATPBAANT (19j3qnop) OA1}BI0 NT NOIL -Vord1dow 8,NVAUG OATFIGOT) OATPIBOT| OAIZISOG] AITISOg 9A19BI0N od od od od OAIQLAON | OATIZBZON | OAIVBZINT od od od od °d od OATIIBOT od od od 9A17R30N od od od OAI}ISOg| 9AI}ISOg (yard) GATPBOT! 9ATPISOT|[NJIQnody od OATVEBON od od od od BATFBBIN | OAIVBIAN| OATVUAON, od od OATPISOT| [NJZQnod od OATPISOg od od od PATZBAIN | BATFVBON | OAIqUVZON, od @ATFIBOd] OATPSOG| OATZISOg (yutd) | (amo1q) OATpIsOg | [NJIqQNOd| aaryuIeN od @ATyudoN od od od od BATPBION | OATFVAON | OATVBSIN (jurd) (pa) (urd) od INj3qnoq | aatytsog |[Nj3qQnoq OATFISOg od OATIUBONy | OATYBDON, (Moy aa (par (amoiq Aqatp Aso) 44 at) quvysur) OAT}ISOg od SATPEBON | VAIQUAN (urd) OATHSOg | OATBAON| OATPBAONY | OATYVBONT (sinoy ns (simoq > #o 0} SoyNurUl QT) ¥¢ 07 So4NUIUT (QT) By CO a (NOLLO108 -IWIGOW 8 NVALYUVA UAHLA) NOILVO -IMIdOW §,NQHdTVH OATIISOg od OATISOT od ““uouuBYys "Ty WZ Iny7qnoq} oArztsog] [Ny3qQnoq| SAIyIs0g|*****‘eou0IMET “gq “N 9ATZBBZON | OAIZUAAN | OATQBAON | GATVBFON | **'''* prsqqny "sg *M BAIJISOg| SATIISOG od od : AQ|SBAD "OM SATZBAON | [NJIGNO| SABO) PATPSOq]" ““SPITGO “V¥ “N (qurd “a5) od og Tnyqqnoqd] earyedon | souTaB AA “oT od OATFIBO od oa oe Une “¥ ) pus yeapoyY sniyne (par 10 G1dWvs (qurey) OAIZIBOg| [NJIQnNO0d| oAIVIBOg] OATpISOg]*"'*'****UOMUBYY "TT “YW od od od Gat} Po2ee9 OOUdIMB'T “ “N SATFBBON | OAIVVSON| OAIVVSON | OAIVBAON | °°" pawqqny] “S “MM (umoiq) [Npqnod] [njyqQnoq| osansog od “KORA "OM od od od 1ny3qnoq |* “SPITGO “VN OAIPBAON | OATQVAON | GATFBBON| OATIGBAON |° *OuUNIBA “OO OATHEOT) OAIZISOd| OATPIBOG) OAIPIBOq|" United “¥ ')) pus yoaqyoy sniyne : G1dWvs (urd) (yard) Iny3qnod od Inyzqnod od SOOT S eileen od od od od “OOUDIMVT “F “N od od OATZERON od “““prsqqnH *S "AM pol od od Aq od Aa[BVOYN "OM °d °d od od i | APIMD SVieN od od od Yates heer 20 JOUIIBM ‘EO DAIFBAON | OAIPVSON] | AATZBZONT | OaTpwAONY | ume “VD pue yoaqoyy sniyne tv G1dNv8 (sanoy $2 0} BeyNUTUT OT) (sunoq (aIOV O1N0THO #Z 0} So9nurut QT) ISX1VNV -OUdAH INGO UGd QZ) NOWvowt -10OW §,LOUVHNIGY isa. IVNIDINO ce sic arta ANY G1dANVS *hawoy fo sajdups xis uo sz)nsaL aayn.toqn}0,) ‘SUOLVUOUVTTION AO SLTOSHY 477 SACCHARINE PRODUCTS SHANNON: 1915] *q oq 0} poumnsse J YOIYM o[duive & UO O18 YIIMOIOY poysodas 64]NseI ayy ‘a1OJoION,T, OAT}ISOg og used daaq od uoaid deaq od od og og oq od oq od od od OATIIBOG (MoT]a4 eyed) od OATPISOg (MoT[a4 9311) OAI}IS0g] SATZISOG (qaj01A en{q) OAtVISOg| Od OATYISOG] PATHISOT OAlzEZaN | [NjJZQnOG od od od od od od od od od od od oq od oq od od °d od @ATHISOg (dat1r 4u18y) BATA TSO”) (urd) BATPISOT BATPISOg (furs Furey) OATPISOg od OATPISOT ({nJ3qQnop) BAT}ISOG OC oa od od DATIIBOg yurd OATHISOT @AT}ISOg OAI}B90 NT od OATFISOT SATVBDONT °d od °a °d od tat od od Cat °d a °d eAtyIsog] od OATIVBIN| OATIISOT SATISOg] AATPBIIN BAIyBSAN | [NJZQNoG Oo” at od OATPIBOg od OAT}BION Cla Cla od od OATTISOg og 9ATPBB9N od od od od od od °d °d od od AATHISOg OAT} BION PATPISOg od BAIZBAAN Cat od °a °d Mal oAtqIBOg SATPBIONT od ‘a]GWIts S1g} UO JUCTIdIYs SULINpP pesuvIIeEIP OBIE JEqR] OU. 1 SATIISOg OATPBION od Tials(ateiels uouusyg “Ty “uy on00s9 aoueIMey ‘g ‘N ee PACA OE Sa “s “ATSBAD “OM SERIO: View sores JOUTIBM “HO Santee unt “V"O pus yeapoy sniyne i G1dWvs8 pooedong uouunyg “TW sfeiofefels ooueIMEy “gq ‘N pee paeqqnH ‘S “At use neespere ais) aayONy Tress SPITTO "VN “oobeeane et aE uryng “yD pus jeayoyy sniyne 10 G1dWvs best SPITGO “VN JOULLIVM “AO Wt EL far) pus yeazioyy sniyne id G1dWvs 478 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. 1, No. 3 Conclusions as to purity of samples. NRO ee Sa — See | ee | a Julius Hortvet and G. A. * | | : Parkineeeo ste scscnesce = Pure | Adulterated | Adulterated | Adulterated | Adulterated | Adulterated. C. E. Warriner. Bead Glcnemcnonontc aor odccan cus MUaesmEcccoed harercsccanadd bRaasboaticne 34 N. A. Childs... W. C. Geagley. Do Doubtful i) | fe) i) W. S. Hubbard Do Pure Pure Do | Do | Do N. B. Lawrence... wjeeiei| folate ey oteefes|feretetetetareretoletaieie Petuleistsietetatststaistota ftaia’ ale fatelatataletaferats ber tietetni=[e/otatatetaletal stale islet teammate F. L. Shannon............| Pure | Adulterated | Adulterated | Adulterated | Adulterated | Adulterated \ COMMENTS OF COLLABORATORS. Hortvet and Parkin: Test 1.—Not very characteristic when invert sugar is present in small amounts. The color that the reagent takes on serves to hide the results. Test 2.—The reagent must be freshly prepared and when so prepared seems more sensitive to small amounts of invert sugar than reagent 1 Test 3.—A delicate test, but the intensity of the reaction is about the same on Samples B, C, D, and EH, although somewhat greater on Sample F. Test 4.—Offers a variety of carmine shades, which confuses one when deciding on the amount of invert sugar present. The test requires immediate observation. Test §.—Much trouble was experienced with the formation of an ether emulsion. The test did not give as decisive result as 7, 8, or 9. Test 6—Gave very poor results, no ring coloration developed for about five min- utes and then only a faint and unsatisfactory result. Test 7.—Seems to be a more conclusive and satisfactory result than any of the first six. The intensity of the color would seem to correspond closely with the per cent of invert sugar present. Test 8.—Seems to be a good test. The shades of yellow vary with amount of in- vert sugar present. Test 9.—Appears to be very sensitive to amount of invert sugar present. N. A. Childs: From the above tests, it is my opinion that A isasample of pure honey; B probably contains a very small percentage of adulterant, D, C, E, and F are adulterated, the percentage of adulteration being much higher in D and F than in C and E. W. C. Geagley: Sample B gave results that were questionable in all tests but Halphen’s, Hartman’s and Feder’s, in which cases the reaction was clear and dis- tinct. In all other tests there would be a question as to the purity of this sample since pure honey sometimes reacts slightly. Sample F contains the largest amount of invert sugar, C next, E next, and D next; B contains a very small amount, if any. Aispurehoney. In Bryan’s modification of Fiehe’s test, I find that an emul- sion forms when ether is added to the 50 per cent honey solution and shaken that will not break up. By shaking gently this can be overcome to some extent. W.S. Hubbard: I considered D, E and F to be adulterated and prefer Hartman’s modification of Fiehe’s test. DISCUSSION. It is quite evident from the reports that artificial vert sugar can readily be detected by most workers when it is present in amounts of 20 per cent or over. When the amount is only 5 per cent, however,some difficulty is experienced. It also seems that the nature of the adulterant has something to do with the possibility of its detection. But very little 1915] SHANNON: SACCHARINE PRODUCTS 479 difficulty was experienced in detecting the 5 per cent Herzfeld sirup in Sample E, while the majority were doubtful as to the presence of an adulterant in Sample B, although they both contained only 5 per cent of the adulterant. This, I believe, can be explained by the fact that the Herzfeld sirup was made by heating, while the commercial invert sugar was no doubt made in the cold. This fact would also explain the discrep- ancy in the reports where the depth of color has been used as a basis of judgment for the amount of artificial invert sugar present. It would seem that one cannot use the depth of color as a guide to the amount of adulterant present. Comparing the original Fiehe’s reaction with the various modifications, the only advantage gained by any of the modifications seems to be in the keeping qualities of the reagent. The original Fiehe’s reagent turns dark on standing and masks the reaction, although with some experience this feature does not seem to interfere. Some workers have recommended that Fiehe’s reaction should be permanent after 24 hours in order to draw positive conclusions, although it must be remembered that samples of pure honey will give a pink coloration on standing. This appears to be a good recommendation, as three obtained positive reactions in Sample B. Bryan’s modification of Fiehe’s test seems to have been quite satis- factory in the hands of the collaborators. Three report a positive reaction on Sample B, one doubtful, and three negative. Some difficulty is ex- perienced with the ether emulsion in this test. Hartman’s reaction must be observed immediately to be of any value as a pink color develops in 10 minutes and cherry red in 24 hours on pure honey. Brown’s anilin acetate test gave good results although this reaction cannot be depended upon when the adulterant is less than 5 per cent. Four report negative results with this test on Sample B, while three report doubtful results. Feder’s anilin chlorid test develops a faint pink color on pure samples, although after standing for sometime this disappears. The red color developed in the presence of invert sugar is very pronounced. Three report positive with this test, one doubtful and three negative on Sample B. The benzidin acetate test and the B-Naphthol test do not seem to have been very satisfactory when the adulterant was only 5 per cent. But one reported a positive reaction with the former test and only two with the latter. RECOMMENDATIONS. It is reeommended— That the work be continued for another year, studying the tests which proved the most satisfactory in the hands of all the collaborators, (1) 480 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. 1, No. 3 The original Fiehe test allowing the color to develop for 24 hours; (2) Hartman’s modification of the Fiehe test, noting the color immediately after the addition of the reagent; (8) Bryan’s modification of Fiehe’s test, changing the word “‘vigorously”’ to “gently;” (4) Feder’s anilin chlorid test: with a view to adopting these as provisional methods in 1915. REPORT ON FRUITS AND FRUIT PRODUCTS. By H. C. Gore (Bureau of Chemistry, Washington, D. C.), Associate Referee. ESTIMATION OF MALIC ACID IN FRUIT PRODUCTS. The work reported last year formed the ground work for further de- velopment of the uranyl acetate method of estimation of malic and tar- taric acids, and indicated that a new polarimetric method for the esti- mation of the two acids, using ammonium molybdate instead of uranyl acetate as the activating agent, was in sight. In addition to the conditions necessary for the successful use of uranyl acetate laid down by Yoder, and those for use of ammonium mo'ybdate indicated by the writer, L. W. Andrews found the time of standing after mixing to be an important factor, and that exposure to daylight must be avoided particularly in case of tartaric acid. In the case of malic acid it is possible to state a general method widely applicable to fruit products and other substances. Two reagents are used in separate solution. Uranyl acetate gives polarizations to the left, ammonium molybdate to the right. Well-agreeing results are usually obtained. Estimation of tartaric acid using the named reagents has not been worked out. The situation is here more complicated than in the case of malic acid. GENERAL METHOD FOR MALIC ACID APPLICABLE TO FRUIT JUICES, JELLIES, SIRUPS, AND OTHER FRUIT PRODUCTS, PROBABLY INCLUDING VINEGARS AND WINES. Determine the amount of free acid present by titrating a portion of the sample with standard alkali and calculate the weight of solid barium hydroxid required to nearly neutralize the free acid. Add to 25 ec. or 25 grams of sample 2} to 3 volumes of 95 per cent alcohol; this will precipitate the pectins. With jellies and heavy sirups special precautions have to be observed to prevent loss of malic acid due to imperfect solution of the sample in alcohol or to imperfect washing. Filter off the precipitate with suction and wash well with 95 per cent alcohol. To the filtrate add sufficient powdered barium hydroxid to nearly neutralize the acidity. Stir until reaction is complete 1 Read by M. G. Mastin. 1915[ GORE: FRUITS AND FRUIT PRODUCTS 481 and then add three to five drops of an aqueous barium acetate solution containing 50 grams in 100 ce. or more if required, for the purpose of providing an excess of barium. Make up the solution to about 375 cc. with 95 per cent alcohol and digest on the steam bath until the precipitate settles readily after being stirred. Filter with suction in a cup-shaped filter and wash well. Dry the precipitate thoroughly, digest with hot water and make up after cooling to 100 ce. This amount of water is sufficient to dissolve barium malate up to amounts as large as approximately 0.9 gram per 100 ec. Filter and treat separate portions of 25 ec. each as described later. To this point the method is adapted from part of a method given in a preceding report,! there designated as the Yoder method. In its present form it is the result of the work of M. G. Mastin, of the Bureau of Chemistry. He finds this procedure applicable to fruit products containing considerable coloring matter and even to those containing tartaric acid. Where citric acid is present in addition to malic acid, as in orange juice, the results were somewhat higher by the molybdate than by the uranyl acetate method. URANYL ACETATE PROCEDURE. Prepare a solution of uranyl acetate, UO2(C2H3;02)2.2H.O, containing 33.966 grams per liter of the pure salt. The uranium content of the salt should be determined (by ignition to U;Os), and if it is a little below the theory (56.18 per cent), a cor- respondingly greater amount of the acetate is to be taken. If it is several per cent too low, a purer material should be obtained. Mix 25 cc. of this solution with 25 ec. of the solution of malic acid to be tested, which should be neutral or faintly acid, and should not contain more than 1 gram in 100 ec. of malic acid. Keep in the dark for at least five or preferably ten hours before the final readings are made. Polarize in a 200 mm. tube at or near 20°C., using white light and the bichromate cell. Table 1 gives the concentration of malic acid in the solution polarized: TABLE 1. Grams of malic acid per 100 cc. corresponding to optical rotatory power in Ventzke degrees, in presence oj uranyl acetate. Ate MALIC ACID DIFFERENCE AVE MALIC ACID DIFFERENCE 200 mm. grams per 100 cc. 200 mm. grams per 100 cc. 1 0.039 9 0.316 : 0.035 0.034 2 0.075 10 0.350 0.035 0.034 3 0.110 11 0.384 0.035 0.034 4 0.145 12 0.418 0.035 0.034 5 0.179 13 0.452 0.034 0.034 6 0.214 14 0.486 0.034 0.034 7 0.248 15 0.519 0.034 0.034 8 0.282 0.034 1 Bur. Chem. Bul. 162, pp. 65-66. 482 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 3 AMMONIUM MOLYBDATE PROCEDURE. Prepare a solution of ammonium heptomolybdate, (NH,);Mo7;02;.4H2O, contain- ing 172.8 grams per liter of molybdenum trioxid. For this 212 grams of the pure salt are sufficient, but more or less of the salt at hand may be required according to purity and content of water of crystallization. Also prepare a solution containing 300 grams of acetic acid per liter. Mix 20 ce. of the above solution of ammonium molybdate with 25 ce. of the solution of neutralized malic acid to be examined, in a 50 ce. flask, make up the volume with the solution of acetic acid, and allow to stand in the dark for about 4 hours. Polar- ize as directed above. Table 2, worked out by Andrews, gives the concentration of malic acid corresponding to the polarization: TABLE 2. Rotations of malic acid in ammonium molybdate solutions. °V MALIC ACID DIFFERENCE | °v MALIC ACID DIFFERENCE 200mm. grams per 1000. | 200 mm. arama per 100 ce. 1 0.027 | 21 0.496 2 0.052 Laseteasl 22 0.519 Sati 3 0.077 Pe | 8 0.542 Ceo 4 0.100 oe 24 0.565 aa 5 0.124 | 25 0.588 6 0.148 ate | 26 0.611 oe 7 0.171 Pes | 27 0.634 Bee 8 0.195 ig | 28 0.657 is 9 0.218 Age | 29 0.680 a 0.023 || 0.023 10 0.241 | 30 0.703 i 0.264 ae | 31 0.726 ae 12 0.288 ae 32 0.749 A 13 0.311 ae a 33 0.772 ae 14 0.334 ce 1 34 0.795 ee 15 0.357 ! 35 0.818 16 0.381 ie } 36 0.841 a 17 0.404 I 37 0.864 18 0.427 Ria ! 38 0.886 aa 19 0.450 ae i 39 0.909 ar 20 0.473 ee | 40 0.932 he 0.023 | 0.022 1915 GORE: FRUITS AND FRUIT PRODUCTS Ni 200 mm. 41 MALIC ACID grams per 100 cc. 0.955 0.879 1.001 1.024 on So oC) TABLE 2.—Concluded. DIFFERENCE 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 483 Nf 200 mm. 66 67 MALIC ACID 1 1 1 1 grams per 100 cc. 526 549 572 595 -616 640 DIFFERENCE 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 0.022 This method was sent out with samples to a number of collaborators. A sum- mary of the results of the work of several is given in the following table: 484 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. 1, No. 3 TABLE 3. Determination of malic acid in fruit juices. CIDER PEACH JUICE ANALYST Uranyl Ammonium Uranyl Ammonium acetate molydate acetate molybdate per cent per cent per cent per cent A.W. Broomell, Washington, D.C....... 0.60 0.530 0.74 0.74 0.60 0.530 0.71 0.82 0.60 0.550 0.77 0.82 M.G. Mastin, Washington, D.C......... 0.60 0.616 1.07 1.10 0.60 0.616 1.07 1.10 Soe a olla coco: 1.03 1.09 ESTIMATION OF CITRIC ACID. Up to the present time no wholly satisfactory method of estimating citric acid is known, despite the fact that much attention has been given to the problem. Andrews made a discovery which leads to a new method. At the suggestion of the writer, he tried the effect of the presence of citrates on the polarizations of malic acid in presence of acetic acid and salt of molybdie acid. It was expected that where less molybdate was present than enough to combine with both acids, lowerings of the readings would be observed in consequence of the fact that the molybdenum would be partitioned between the two oxyacids. This, Andrews found to be the case. He found, however, that when more than sufficient molyb- denum was present for both acids, large increases in polarizations were observed over the increases which occurred with malic acid alone. Mastin later finds that this observation is probably specific for citric acid. Sodium salts of succinic, lactic and aconitic acids produced no effect on the molyb- denum malic acid polarizations. This observation of Andrews was promptly developed by him into a method using, however, sodium molyb- date, a somewhat unusual reagent. He found that the polarization in- creases due to citrates did not vary directly with the amounts present but that the variations were, however, regular, and he constructed curves from which the amounts present of citric acid could be determined, using appropriate solutions containing sodium molybdate, sodium malate, and free acetic acid. If the solutions, to be tested already contained malates, the amounts present could be estimated from the polarizations with uranyl acetate, and necessary additional amounts could then be added so that the solutions contain the required amount of malic acid. Mastin is now work- ing out the procedure for estimation of citric acid, using ammonium molybdate. For the assay of citrates and citric acid, 2 very simple pro- cedure can undoubtedly be developed. In fruit products the method will necessarily be more complicated. 1916} HARTMANN: WINE 485 RECOMMENDATIONS. It is reeommended— (1) That the proposed methods for malic acid be provisionally adopted by the association, and that the referee for the coming year be instructed to develop these methods further. (2) That the method for the estimation of citrie acid by the use of a solution containing ammonium molybdate, sodium malate, and acetie acid, be further studied. REPORT ON WINE. By B. G. Hartmann (Bureau of Chemistry Food and Drug Inspection Laboratory, Chicago, Ill.), Associate Referee. This year’s work on wine was devoted to a further study of the methods for determining tartaric acid in wines and grape juices. The work was planned in accordance with the resolutions on the subject adopted at the last meeting of the association. The following instructions to the collaborators show the purpose and scope of the plan: INSTRUCTIONS TO COLLABORATORS. DETERMINATION OF TARTARIC ACID IN PRODUCTS CONTAINING FREE PHOSPHORIC ACID. The three solutions submitted for this year’s work on tartaric acid contain vary- ing amounts of tartaric acid and phosphoric acid. In order that products similar to those found in the market might be had, sugar, alcohol and coloring were added to the solutions. The solutions are marked 1, 2 and 3. It is suggested that the total tartaric acid be determined in each of the solutions, according to (1) The bulletin method, described on page 86 of Bulletin 107, Revised; (2) method proposed by Hartmann and Eoff, one-half neutralized; (3) method proposed by Hartmann and Eoff, completely neutralized. Before starting the work, it is suggested that collaborators read ‘‘Proposed Method for the Determination of Tartaric Acid Content in Wines and Grape Juices,” Bulletin 162, page 71. PRELIMINARY PROCEDURE. Although no precipitate is expected to occur in the solutions, it seems advisable, as a precautionary measure, to filter before proceeding with the determinations. Determining the acidity of the samples —The indicator to be used is prepared by thoroughly mixing in a mortar 0.5 gram of phenolphthalein and 50 grams of sodium sulphate. Into the cavities of a spot plate, place about 0.1 gram of the indicator. Measure 20 ce. of the sample into a 250 cc. beaker and titrate in the cold with tenth- normal sodium hydroxid until a few drops spotted on the indicator give a purple tint. It is essential, in order to get a good, clean end point that during spotting none of the indicator is carried into the solution undergoing titration. Place the indicator to one side of cavity and allow the drops to run into the indicator. It will be found that the red color of the sample does not interfere and that, with a 486 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. 1, No. 3 little practice, a very satisfactory end point is obtained. This indicator is especi- ally adapted to phosphoric acid. It has been found that phosphoric acid solutions may be accurately titrated, giving sodium di-acid phosphate. The collaborators are requested to give special attention to the behavior of this indicator, since it has been found that its use is applicable to a wide range of deeply-colored products. In case difficulty is experienced with this indicator, it is suggested that any other means of arriving at the acidity of the samples be used. Express the acidity in terms of tartaric acid per 100 ce. 1 ec. N/10 NaOH = 0.0075 gram tartaric acid. DETERMINATION OF TARTARIC ACID. (1) Bulletin method —Transfer 50 cc. of the solution at 20°C. to a 250 cc. beaker, add 50 ee. of distilled water, and follow procedure described on page 86 of Bulletin 107, Revised, using 20 cc. of alcohol instead of 15 cc. (2) Method proposed by Hartmann and Eoff (one-half neutralized) —Transfer 50 ec. of solution at 20°C. to a 250 cc. beaker; add required amount of normal sodium hydroxid to neutralize one-half of the acidity; add enough water to make 100 ce. Proceed as in bulletin method, omitting the addition of potassium acetate and using 20 ec. of alcohol instead of 15 ee. (3) Method proposed by Hartmann and Eoff (completely neutralized) —Transfer 50 ce. of the solution at 20°C.to a 250 ce. beaker and add required amount of normal sodium hydroxid to just neutralize the acidity of the 50 cc. solution used and add enough distilled water to make 100 ec.; add pure tartaric acid; pulverize the tartaric acid and dry at 100°C. for 2 hours. Keep in a glass stoppered bottle in a desiccator. The amount of tartaric acid to be added is found by multiplying by 0.075 the num- ber of cubie centimeters of normal sodium hydroxid used to neutralize the 50 ec. sample. Weigh this amount exactly. When the tartaric acid is dissolved proceed as in bulletin method, omitting the addition of potassium acetate and using 20 ce. of alcohol instead of 15 ce. Subtract the tartaric acid added from the total tartaric acid found by titration. This gives the actual amount of tartaric acid present in the 50 ec. sample taken. It is, of course, necessary to multiply the figures obtained in the different deter- minations by 2, in order to get the amount of tartaric acid contained in the solutions. Instead of allowing the reaction mixture “‘to stand at room temperature for at least 15 hours,’’ as given in the bulletin method, hold for this time at a temperature not exceeding 15°C. The following is the composition of the three solutions in grams per 100 ce.: DETERMINATION 1 2 3 Acid yas) Cantaricyccen sete ems 4.92 252 1.96 Rotaletartanie;acids\. S.-i eee. 2.00 1.80 1.89 Phosphoniciacidee saeco 1.91 0.48 0.048 Caramel jcatsreceeei sehen oes 1.2 12 itp) Amaranth joc osurschn acer scotia 0.024 0.024 0.024 Canetsugarittacs i. coe cease 5.0 5.0 5.0 Alcohol jis Siac itas dake oes aes 5.0 5.0 5.0 RESULTS OF COLLABORATION. The following are the results obtained by the collaborators on the three solutions: 1915] HARTMANN: WINE 487 Analytical results on tartaric acid in three solutions. TOTAL TARTARIC ACID SOLUTION AND COLLABORATOR nanan ees Method 1 Method 2 Method 3 SOLUTION 1: 4.89 0.70 1.87 1.89 BaG. Hartmanne-- ees 7/20/14 4.91 0.78 1.88 1.90 4.91 0.73 1.89 1.90 en 0.64 1.67 1.82 Bs FB eit ypc ero dtc g/i2/14 | 4447 0.69 1.69 1.82 0.68 1.71 1.87 1.76 18,2000 eee 8/24/14 | 4.80 { Nea me | 1.76 = 1.77 (4.71 0.63 1.66 1.68 me wment: (0.0... 2. 0s: 8/26/14 eas 0.61 1.66 1,67 4.71 0.56 1.67 1.67 Med ingles... ..:<..-: 10/30/14 | 5.06 0 1.61 1.65 SOLUTION 2: 2.57 1.02 1.76 1.75 B. G. Hartmann......... 7/20/14) 42.57 1.02 1.75 1.77 2.55 1.04 1.76 1.75 as 0.89 1.63 1.66 HRMHe Beary cccst-o. secu: 8/12/14 Ee 0.91 1.63 1.69 0.91 1.61 1.69 1.61 : ; 0.67 1.50 Tae 8/24/14| 2.47 tac ie | 1.63 1.60 2.45 Aen Wan |) (less Bs Wert, Saseee 6 6: 8/26/14 | 42.45 { eee 1 aa ne 2 44 1.58 M. J. Ingle.........<.-..| 10/30/14 | 2.54 0.65 151 1.63 SOLUTION 3: 2.00 1.21 1.87 1.86 B. G. Hartmann......... 7/20/14 1.99 1.23 1.86 1.87 1.99 124 1.87 1.88 a 1.19 1.77 1.76 Bis Els Bett yc. eae 8/12/14 ee 1.19 1.77 1.77 1.10 1.76 1.78 F 1.78 f1.05 1.68 MARGE s.-. a eres 8/24/14 1.93 | | o'95 iS 1.77 1.77 1.92 1.22 1.72 1.71 RG) Kent...) ee 9/26/14 1.91 1.21 1.72 171 1.92 1.22 171 1.71 Me J tingle’ 2<).c kee 10/30/14 1.95 0.92 1.63 1.67 488 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. 1, No. 3 NOTES BY THE COLLABORATORS. E.H. Berry—t found it necessary to break the filter paper up as much as possible before titrating the tartaric acid as the paper holds the acid. The indicator was found to be very satisfactory. In highly-colored solutions where with phenol- phthalein solution it is practically impossible to detect the end point, with the indi- eator used in this work, it is possible to detect it to within 0.1 or 0.2 of a cubic centi- meter of tenth-normal alkali. B. G. Hartmann.—The indicator proposed for this work is very satisfactory. It was found that the addition of a few cubic centimeters of alcohol to the mixture undergoing titration helps materially. The alcohol allows a better and more uni- form penetration into the indicator during spotting. Also it was found that if the indicator is too fine, the penetrating power is decreased to such an extent that great difficulty is experienced in spotting. M. J. Ingle.—For ordinary routine work the half-neutralized method seems to be practicable enough, especially in cases where the addition of tartaric acid is not needed to help the precipitation. A procedure found to be of help in making the end reaction on the tile plate more easily visible, was to put an equal amount of the anhydrous sodium sulphate alongside of the indicator in the same depression. Then, by allowing the drop of solution to wet both portions of the powder, the contact between the sulphate blank and the indicator renders it more sensitive. To those having difficulty in using an inside indicator in the dyed-bitartrate solution, it was found practicable to concentrate the color from this solution on wool intro- duced into it. By previously boiling and washing this wool any acidity present is eliminated. Then by allowing the aqueous solution of the bitartrate precipitate to stand on the water bath for an hour or so with an excess of the washed wool, the color is removed. Titrate with the wool in the beaker. R. C. Kent—The indicator is very good for titrating highly colored solutions. The end point is easily distinguished with 0.1 ce. tenth-normal sodium hydroxid. DISCUSSION. The reports by the various collaborators show that the bulletin method is entirely unreliable and useless in cases where free phosphoric acid is present. Two of the collaborators report no tartaric acid in Solution 1 when the bulletin method was used. This is undoubtedly the true result obtained by adhering literally to the instructions issued. The bulletin method calls for ‘one minute of stirring to start crystallization of potas- sium acid tartrate. In a solution containing a high amount of free phos- phorie acid, this time of stirring is too short to start precipitation. That the other collaborators found tartaric acid in Solution 1 is due to the fact that longer time of stirring was used. Methods 2 and 3 (Hartmann and Eoff, one-half neutralized and com- pletely neutralized) although giving fairly good results, are far from satisfactory. That these methods did not give better results is due to the well known fact that alcohol and tartaric acid combine to form esters and this esterification is further augmented by the phosphoric acid content which acts as a catalizer. The esters form rapidly and increase with time, a 1915] HARTMANN: WINE 489 as will be seen from the report, causing a gradual decrease in tartaric acid found, from 0.2 to 0.3 gram in 3 months. For the purpose of saponifying the esters, 50 ce. of each of the solutions were neutralized with sodium hydroxid and 3 ce. of normal sodium hydro- oxid added in excess, and the solution allowed to stand overnight. The volume was then made up to 100 ec. with water, the required amount of tartaric acid added, and the determination proceeded with as described under Method 3. This procedure gave the following results: Total tartaric acid (grams per 100 cc.) i J. R. HOFF SOLUTION CONTAINED fh ege/ik 10/17/14 10/26/14 2.00 1 2.00 2.00 2.05 | 2.08 2.00 (2.09 2 1.80 lth 1.83 | Tees 2 ee 8 1.82 1.88 oe 3 1.89 1.89 1.92 f es 1.89 Sa From these results it appears that in order to arrive at the true total tartaric acid content in alcoholic mixtures, saponification of esters must precede the final determination. In order to verify these findings, it was suggested to J. R. Eoff to apply the procedure to the solutions. His results agree very well with the results obtained by the association referee and are given in the above table. The reports on the behavior of the proposed indicator are very encourag- ing. From experiments made with it, there can be no doubt that its use is of unlimited application in work on highly-colored organic acid solutions. Experiments with the indicator on phosphoric acid solutions of known content gave excellent results, the end point being very sharp when the solution undergoing titration was kept cold. RECOMMENDATIONS. It is recommended— ; (1) That the following method for determining the total tartaric acid content in solutions containing free mineral acids or alcohol be subjected to further study: Neutralize the acidity with sodium hydroxid and add 8 ce. of normal sodium hydroxid in excess. Allow to stand overnight and add the required amount of tartaric acid (0.075 gram for every cubic centi- meter of normal sodium hydroxid added). After the tartaric acid has 490 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. 1, No. 3 dissolved, add 2 ce. of glacial acetic acid and 15 grams of potassium chlorid. When the salt has gone into solution, add 20 ce. of alcohol and stir until precipitation sets in. Allow to stand overnight at a temper- ature not to exceed 15°C. Collect the crystals and titrate in the usual manner. (2) That the indicator described be tried on other food products. MARASCHINO. By J. G. Rinny anp A. L. Subiivan. (Bureau of Chemistry Food and Drug Inspection Laboratory, Boston, Mass.).1 The world-famed cordial, maraschino, was first manufactured commerci- ally early in the eighteenth century in Zara, Dalmatia, from the marasca cherry, a small variety of the European wild cherry native to the Dalmatian mountains. The manufacture of this cordial has continued to the pres- ent day and large quantities of maraschino are still shipped from Dal- matia. The superior excellence of maraschino led to the manufacture of similar cordials in the countries of Italy, France, Holland, and America. The purpose of this paper is to set forth analyses of ten samples of genu- ine maraschino, representing the products of six manufacturers, obtained through the courtesy of the American Consul at Trieste, Austria, and Mr. Nicolo Luxardo. Analyses of commerical samples of maraschino manufactured in Holland, France, and the United States are also tabulated. METHODS OF MANUFACTURE. In Dalmatia during the month of June, marasca cherries are gathered and shipped to Zara. For the manufacture of the best grade maraschino the cherries are pitted, crushed, and allowed to ferment for 4 or 5 days with a small quantity of leaves from the marasca cherry tree; from 10 to 15 per cent pure alcohol is then added to arrest fermentation and to pre- vent the development of wild yeasts and bacteria. One of the objects of adding alcohol to the fermented cherries is to enable the manufacturer to distill the product at his leisure throughout the year. If the fermented cherries are allowed to stand any length of time there is danger of serious deterioration in the flavorand aroma of the product, especial y when alcohol has not been added. The fermented cherries do not yield sufficient alco- hol for proper preservation of the mass. Simple pot stills are used exclusively in the distillation of maraschino spirit and these in most cases are heated by direct fire, although at the present time the use of stills heated by steam coils is being introduced. The type of the still, however, remains practically the same as the original pot still. The first and last portions of the distillate are rejected for the 1 Read by C. S. Brinton. 1915] RILEY AND SULLIVAN: MARASCHINO 491 best grades of maraschino, and a portion of a distillate coming over at about 140 proof collected. The strong alcoholic distillate is stored either in glass-lined barrels or cisterns, or in barrels which have been treated so that the spirit will not extract any color from them. The aim of the manufacturer is to age the distillate when possible for from two to three years. The maraschino cordial as found on the market is made by dilut- ing a certain amount of the strong maraschino spirit with sirup. There is some question as to whether any flavoring materials other than the cherries and leaves are used. The best manufacturers claim to use no artificial flavor. Lower grades of maraschino liqueur are produced from cherries which are more or less unsound and in some eases the pits are not removed so that the distillate may show appreciable traces of hydrocyanic acid. It is claimed by the manufacturers of the genuine Dalmatian maraschino that the best product is made from the wild marasca cherry. If the cherry is transplanted to other localities and countries and cultivated it will not yield upon distillation a product having the flavor of the original fruit. In France so-called maraschino is made by various methods, which may briefly be classified under three heads: (1) The cherries are crushed and allowed to undergo alcoholic fermen- tation in the presence of a certain amount of the cherry leaves. After the fermentation the product is distilled and either a very strong spirit known as marasca spirit containing 40 to 50 per cent alcohol collected, or the fermented cherries are distilled in such a manner that a dilute spirit, 8 to 15 per cent alcohol strength, is obtained. This is called eau de marasque or marasca water. (2) A mixture of black cherries, raspberries, or other fruit and cherry leaves, with a small amount of peach kernels and iris is fermented and disti.led and a strong distillate obtained which is used for the manufacture of the cordial. (3) Essences of peach kernels, orange flowers, jasmine, and vanilla are mixed with pure alcohol and an artificial spirit obtained which is later made into a cordial. The method described under (1) is generally similar to that followed in Dalmatia. It is claimed that the cherries used are of the same variety as the original marasca cherries and that these cherries grow in Italy, Greece, and France as well as in Dalmatia. From information obtained from various sources it appears that it is well recognized in France that the marasca spirit or marasca water obtained from the native wild cherry is distinctly inferior in flavoring strength and quality to that produced in Dalmatia. Information from similar sources makes it evident that genuine marasca distillate from Dalmatia is often claimed to be used by French manufacturers. 492. ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 3 i In Holland so-called maraschino has been manufactured for many years; the following statements were made by a Dutch manufacturer: “Tn the trade, the term ‘Maraschino’ means a liqueur produced by the distillation of the kernel of stone fruit, generically the Prunus acidus; it may be simply the cherry, or the May Cherry, the black cherry, Morello, or Marasque. It is said that this general variety of cherry originated in the eastern and southern countries of Europe where the Marasque kind has predominated. “It is believed that in the beginning, over a century ago, the Marasque was the sole or chief variety of cherry from which Maraschino was made. But in course of time, it is related, to suit the public taste, this liqueur was distilled by producers all over Europe, from other varieties of the cherry as well as the Marasque—some- times blending Marasque and other kinds, sometimes using no Marasque what- ever. Sometimes, also, other substances were added, as flavoring, to please the consumer. All this time the liqueur was called Maraschino, and thus this became a generic term, without specific reference to the Marasque or Marasea cherry. ““At the present time, as appears from the best information obtainable, no maker of Maraschino in this country uses cherries brought from Dalmatia, but the makers do use local or other varieties as near like them as possible. For instance, the X firm inform me that they use cherries grown in this country from real Marasca sprouts which they import and plant here. “The member of the firm of X says that the flavor of his Maraschino is reenforeed by other substances * * * these substances are a trade secret which he could not divulge.’’ The following table gives the analysis of ten samples of genuine Dal- matian maraschino, nine samples of the French product, four samples of the Dutch product, and three of the American; also a composite analysis of Kirschwasser taken from Kénig, volume 1, page 1514. Description of Samples Analyzed. 2216-KK to 2221-K. Characteristic flavor and aroma of true maraschino. Slight suggestion of Kirsch. 2222-K. Weak flavored, no maraschino flavor; very little, if any, cherry distil- late; test for hydrocyanic acid not regarded as conclusive. 2223-K. Cherry kernel flavor; benzaldehyde odor noticeable on diluted sample. 2227-K and 8. F. 3249. Flavor very weak, possibly derived from wild cherries. NY. 38512. Nearly all spirit, with a slight flavor of maraschino. NY. 38513. Spirits flavored (rose and syringa suspected); consular report shows that in district where sample was made alcohol and artificial flavors are used with either Zara marasca water, or same from Grasse district, France. NY. 38752. Weak flavored, may contain a small amount of maraschino. NY. 39752. Does not have flavor of maraschino; may contain a cherry distillate; benzaldehyde suspected by odor and taste; manufacturer admitted later that sample was not prepared from marasca cherries. 2224-K and NY. 26099. Perfumed odor rose present. 2225-K and NY. 26047. Artificial flavor present; no maraschino flavor; see description of Dutch maraschino. 2226-IKK. No maraschino flavor; benzaldehyde suspected by odor and taste; made from cherries, pits, alcohol, ete. 3550-H. No flavor of maraschino. 1687-K. Has maraschino flavor; use of imported marasca distillate suspected. The analysis of genuine Dalmatian maraschino shows it to be an alco- holie cordial containing from 30 to 44 per cent of alcohol, and 26 to 36 per 493 MARASCHINO SULLIVAN RILEY AND 1915) “UBATIINS “IT “VW Aq §16I-606T pezAyeuy 1 veteeeeeeeeeeeel pegge 2°29 0 6F P18 gop foc vecseeeloosoprep ed "qa ‘Soy :UGdSSV MHOSUIH od ows} QUIBT oarquae N £798 GLI 00 6'0 om | 6 TSG foot oe] DERE O SIAC» Eiki ¢ aATqsOg AATISOg 06 Ver 9g ia) iad gT 6+ ORGE Ie (OCOR MTaTwed | erates one STEDGOD queselg guasorg B'SZ 1°88 Tel £0 +S IG |p ae) Ola2 1] Sts | hGiquey, » |fPoeecogcne DD Geld INVOIUGWY Uisre terstelsce seit aamwog frrreecrete reese fere seen 6°29 cowry | 9'F 8-29 seeeeel grap Legge [rere t- “5008 CRN quesolg Jueselg 0°91 ¥ 621 “FOI 00 Le £91 Fav, SOF | S2°ee | SFT T i "M-S20 aiatelolereis qureg [occtecreteceeeeefectertee] rege 0:0 80 6°oI Pateee[eeTenteal gouge [ettecnes|ee “166092 “KN OATISOT OATIISOT @ SI e'1s ¥'S€ 00 fs 06 oF 86 | 68C8 | SOLE jt Faas :HOLAG paqoaqap ouoyy [eet oor eee] tng eee: el gg “19 seseeeel gigg [oreeecedleseeeeeeee trey ge ox NT od oq od 00 00 elF | PETS | ‘ “109288 “AN au0N oT0N au0N, 0:0 | 10°0 9IT ‘| ese “STSSE “AN teens eeee quiwy |e: veeet renee 9081}919}T| QT 0°02 ‘| 6°cF “W8F09% “AN au0N ee see au0oN 00 | 88°T 212 6 LF “IG8é “AN og OATQURO NT. ||Poaties asseien es BOI 9 00 6 6F 16FCE “A'S 8AT}BBON 90BI} JUIET [nyaqnoq 00 6°0 6 Les bs arid OAIHSOd aarysog ¥°92 17'0 | 68'T VST £28 S800 9A1q8s0N Sour} FULT ATYOON 00 91 00 fre ES Fla THONGUS t Zz a081L 00 0% 6 F 20 T'93 | $6°0E | QOOT Jot unurToTyy 8°21e 0°29 6¢°0 @ 8% 8° 2cr | $8 698 | 6S FR | OOLT Joo unurxeyy 9A19830N gig | evry, 00 0% for | 26 1S |) Atal eyes SoBly, 8 ZIG F909 soul, @ 8% 8 cel | #9 1oFE | F6°0E | 9OT'T quyey Alo, T 02 Sle 00 $9 F 62 aid 9 8& | 62°SE | 9OT'T °d 8 OIT F'9g 69°0 88 99G | $'8I G1 | 90°th | L90°T OATYEDON, 8 82 £96 co 0 ¥8 8°93 8 °€T 6 FE | Shoe | G60'T oq fee £08 9°E1 an) Le 6°21 Z0 6'oE gee | Sort 90s, PAITS0g re £68 8'£8 £20 8 808 | oor | $82 | 62° FR | L90°T datyezaN, aaryedoN 4 Lz 9°6 00 6L 6F $F G08 | 6z'Ie | 960°. OATHSOT aATHEOg ia S°¥8 0°29 ze'0 ae 66 16 6°62 | €L°c& | T60'T 9A17890 NN qurey AIOA ooBLL Lig Lilt 00 £F Lot | oF 9% | 79 LE | 990°T :ONIHOSVU VIN NVILVWIVd @NIONGO be > n Le} i>] > & > pid ats cc] 3503 9803 Be |EESES = |gee| ge | oe | 338] 5s = 145 7o > a n ES = ayvuBdooydyng | 1addoo ovvensy (yoord 35 ze Sto [Bins Any & Sieg ary are 5 mh o00T'000'001 s°0 |REELS a © p ox Or 2 AIGIUIVA dod s31q) Po |-ae 3 o B e2 a oD 2 a ad kHad1vzNag BST aw fe 10)| 1305201] 26" 201i eee pera 5.70} 2.50 |Trace | 46.40 | 43.00 | 40.54 Peppermint........ Onl Ze GON | eeer ere Pactra: Epa ia cKO) 26 100 nie SO) paneer 5380))- as ars || 46 00) |e ae 44.00 | 42.20 | 40.20 AA 20))|42:54.0) 18a crete Nite 4.80] 2.40 | 0.20 Loree) Meester Iso biota fi nema kee ; 2h P20 b oireratr ene ttevocsiere el crseane eo Awe ivullieciaedllce sue 54.80 | 52.50 | 50.60 | i eeeet| leicester _ ee ra ep Trace 24.60 | 21.70 2Trace aan 42.40 | 40.30 PEarMINte.... 2... é SAO Sen oe 4.49 |62.30|...... 5 eee ree AS OO ented | tree | °5.10 | £2.60} 50.54 Wintergreen’...... 1 a 2.10 | None ; SAO ea Brsccr al tehess ee pd “2.50 | “0.50 eOUl. we eee fe twee or, aa ea) CC Rose (85.5 per cent alcohol—a stand-| _ 40.40 ; ard brand)...... None) Peer ih: |ie dae SPADA RG ore bes | aon 3 OMONr | ter aalhietnnt 1 The percentages of oil present in all extracts except the rose were 5 per cent, 2.5 per cent and 0.5 per cent. 2 Carbon bisulphid method exactly as directed. ’ Carbon bisulphid method except that pump was not used at 100°C., and heating here continued but 10 to 15 seconds. 4 Ether extraction, heating at 75°C. with 2 cc. saturated sodium chlorid solution. ° Ether extraction, heating at 50°C. with 2 cc. saturated sodium chlorid solution. 6 Carbon bisulphid, heating 4 minutes at 55°C. only plus sodium chlorid. 7 Winter green by saponification gave 4.85 per cent and 4.85 per cent. solvent at a moderately high temperature. Spearmint oil seems to show a similar tendency. With weak extracts, however, particularly when prepared with strong alcohol, the separation is either very far from complete, or entirely negative. Were larger proportions of sample used and greater dilution practiced than is possible in the ordinary Babcock bottle, the results, as pointed out by Hortvet and by Mitchell, would undoubtedly be very much more satisfactory. Nevertheless, because of the impossibility of uniformly separating all, or nearly all of the oil present, and of the application of any really definite or constant correction, methods based upon precipitation do not at present seem to hold forth much promise, although the principle can, without doubt, be applied advantageously in certain cases. 504 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 8 Carbon bisulphid method.—In addition to securing complete extraction, we have involved in this the rather difficult problem of adjusting the method and tempera- ture of evaporation with such nicety that the solvent will be completely eliminated, on the one hand, without loss of oil, on the other. The tenacity with which small amounts, not only of chloroform and bisulphid, but even of ether, are retained by the oily residue, is rather remarkable. While there can be no question that in the case of most oils used as food flavors, when present in fairly large quantity, extrac- tion of all but unmeasurable traces can be effected by as little as 1 cc. of chloroform or carbon bisulphid, yet when the proportion is less than 1 per cent and particularly when the solvent is strong (95 per cent) aleohol—as in the 1 ee. dilution here involved— the separation is likely to be more or less incomplete. It is, of course, obvious that inasmuch as the latter circumstance would almost never be encountered in practice, this objection is not quite so vital as might at first appear. Extract of rose is, however, a case in point, the sample included in these tests containing 85 per cent of alcohol, with less than 0.5 per cent of oil. The writer trusts he may not seem lacking in a proper sense of modesty if he ventures to point out that the ether extraction method, as described below, proved to be the only one of the three volu- metric procedures capable of properly working this particular case. Objection may be raised to the provision concerning heating in a boiling water bath under diminished pressure. While this may involve no serious loss in the case of a few of the oils, there seems to be but little room for doubt that a serious loss may occur with others under this condition. Infact, our work would seem to indicate that even at 70°C., with the pump, material volatilization of oils of the character of nutmeg takes place following expulsion of the bulk of the solvent. In reality, the writer is convinced that the extended use of such high temperatures is unnecessary, although in some instances they may do no harm. Finally, it might be well to investigate the possible objectionableness of the addition of as much as 1 ce. of strong hydrochloric acid in the case of such extracts as cassia, cinnamon and clove. In connection with earlier work along these lines, the impression was gained that such causes a tendency to low results through de- composition. On the other hand, the writer is still unconvinced as to the objection- ableness of the simpler 1:2 sulphuric acid floating medium for oil of wintergreen, provided such is cautiously added to the well cooled oil separate. Ether extraction method.—This procedure was proposed by the writer in 1911 (J. Ind. Eng. Chem., 1911, 3: 252) as an improvement over his original chloroform method, concerning which certain criticisms had been raised. While it is believed that the present carbon bisulphid modification is capable, possibly following a little further adjustment, of affording good results in a rather large range of cases, yet the writer is convinced of the general superiority of the ether method, believing not only that it is applicable to a greater variety of extracts and oily preparations generally, but to the most exacting conditions as regards strength of alcohol and proportion of oil. Under the present practice in this laboratory the temperature of evaporation is such that no serious loss is likely to occur in the case of turpentine and oils of similar volatility. The principle involved in the original chloroform method, that practi- cally complete extraction can be effected by means ofa very small volume of solvent. has been lately further extended in the ether process, through the elimination of the third shake out. While not quite so rapid as the bisulphid method, yet it is not a lengthy procedure. Following are the details as at present observed: Pipette 10 cc. of extract intoa4 ounce separatory, the stem of whichis cut off short at the stopcock. Add 50 cc. of cold water and, except in the cases of cinnamon, 1915) PAUL: FLAVORING EXTRACTS 505 cassia and clove, about 0.5 cc. of strong hydrochloric acid. Shake with 15 ec. of ether, draw off aqueous layer and transfer ether extract to a small-mouthed 50 ee. flask. Again shake with 10 cc. ether, reject aqueous portion, and wash the com- bined extracts with 10 cc. of ether-saturated water. Transfer the extract by means of a small funnel to a Babcock milk bottle graduated in tenths, and add 2 cc. of satu- rated salt solution. Connect stem of bottle with vacuum pump and immerse in a bath at 50°C. Holding bottle at an angle of 45°, shake continuously with a rotary motion (at first gently), until all but a small residue of ether is eliminated, which, with a good pump, will require 2 to 3 minutes only. Continue the heating for about 2 minutes beyond the point when, on removing bottle from the bath about every 15 seconds and giving contents a vigorous snap, no ether foam is observed. Remove from the pump and test completeness of ether removal by quietly immersing from 2 seconds in a boiling water bath; on the instant of removal simultaneously shake and apply a test flame. Repeat, and if not found ether-free, return to the first bath for another minute. Care should be observed not to expose the nearly ether-free oil to the temperature of boiling water for more than a very brief period, nor must the evaporation under the pump at 50° be unduly prolonged, else loss will result. Finally, cool, add salt solution and centrifuge for 1 minute. While, using 10 cc. of sample, it has been found that one extraction only of 15 cc. ether is usually sufficient for the removal of all but unmeasurable quantities of oil when the latter is present in considerable amounts, yet the second extraction is necessary with small amounts. In order to reduce the error, it is our practice, when the proportion of oil is much less than 5 per cent, to take 20 cc. of sample in which case a third extraction with 5 cc. of ether is advisable. With most extracts, a clear separation occurs readily, and the use of hydrochloric acid would seem of doubtful necessity. The acid was, however, found very help- ful in the case of the nutmeg and spearmint, both of which extracts tended to give a persistent cloudy water-alcohol layer in the absence of such. The addition of the 2 cc. of brine solution seems to be effective in assisting the elimination of the last traces of ether. RECOMMENDATIONS. In view of the above results, comments and conclusions, I would re- spectfully reeommend— (1) That the saponification method of Hortvet and West for winter- green extract, as described in the Journal of Industrial and Engineering Chemistry, 1909, volume 1, and slightly modified in Bureau of Chemistry Bulletin 152, page 141, by the then associate referee, R. S. Hiltner, be adopted as provisional. Method.—Mix 10 ce. of extract in a 100 cc. beaker with 10 ec. of potassium hy- droxid solution (10 per cent). Heat ona boiling water bath until volume is reduced about one-half. Add a distinct excess of dilute hydrochloric acid, cool and ex- tract with three portions of ether, 40 cc., 30 cc. and 20 cc., respectively. Filter the combined ether extracts through a dry filter into a weighed dish, wash with 10 cc. of ether and evaporate spontaneously. Dry over calcium chlorid in a desiccator and weigh. The weight of salicylic acid thus obtained, multiplied by 9.33, gives the per cent of oil of wintergreen by volume. 506 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 3 (2) That the following method, devised by Hortvet and West, and described in the Journal of Industrial and Engineering Chemistry, 1909, volume 1, number 1, be made provisional for anise and nutmeg extracts. Method.—To 10 ec. of extract in a Babeock milk flask add 1 ce. of hydrochloric acid (1:1), then sufficient half-saturated salt solution previously heated to 60°C., to fill the flask nearly to the neck. Cork and let stand in water at 60°C. for about 15 minutes, occasionally giving the flask a twisting motion, and centrifuge for 10 minutes at about 800 revolutions per minute. Add brine till the oil rises into the neck of the bottle, and again centrifuge for 10 minutes. If the separation is not satisfactory, or the liquid is not clear, cool to about 10°C. and centrifuge for an additional 10 minutes. (3) That the following slight modification of the Howard-Mitchell method, which was studied last year for peppermint extract, be now pro- visionally adopted for peppermint and spearmint extracts and also as an additional method in the case of wintergreen extract. Method.—Pipette 10 cc. of the extract into a Babcock milk bottle, add 1 ec. of carbon bisulphid, mix thoroughly, then add 25 cc. of cold water and 1 cc. of con- centrated hydrochloric acid. Close the mouth of the bottle with the thumb, and shake vigorously, whirl the bottle in a centrifuge for 6 minutes and remove all but 3 or 4 cc. of the supernatant liquid, which should be practically clear, by means of a glass tube of small bore and aspiration. Connect the stem of the bottle with a filter pump, immerse the bottle in water kept at approximately 70°C. for 3 minutes, removing from the bath every 15 minutes and shaking vigorously. Continue in the same manner for 45 seconds using a boiling water bath. Remove from the bath and shake while cooling. Disconnect from the suction and fill the bottle to the neck with saturated salt solution at room temperature, centrifuge for 2 minutes and read the volume of the separated oil from the top of the meniscus. Multiply the reading by 2 to obtain the per cent of oil by volume. In the case of wintergreen, use as floating medium, a mixture of 1 volume of con- centrated sulphuric acid and 3 volumes of saturated sodium sulphate solution. THE RELATIONSHIP BETWEEN THE ALCOHOL-SOLUBLE SOLIDS AND ETHER-SOLUBLE SOLIDS IN STANDARD GINGER EXTRACT. By C. W. Harrison snp A. L. Suntivan (Bureau of Chemistry Food and Drug Inspection Laboratory, Boston, Mass.).! During the course of the regular inspection work in connection with the enforeement of the Food and Drugs Act, it has frequently become necessary to determine the purity and strength of ginger extracts and tinctures. The high cost of alcoho] leads to the attempt to put out ex- tracts of low alcoholic content and of little flavoring strength. The following two analyses are typical of this class of extracts: 1 Read by C. S. Brinton. 1915] HARRISON AND SULLIVAN: STANDARD GINGER EXTRACT 507 Per cent Per cent INT eo CCG) NS ae en Ces ee ote BiG otic ou nodosa conte 46.95 24.30 SONG SS GH aamionteena peta iene te acmoro caged comunbanoOeae 0.60 0.79 Wiater=solubl eisolidsscitancists.5 ce akc eine oe eeiiates Stes 0.77 Hther-solublesolidssansae ones osceisiie cee ee nee eee. 0.06 None Ginger, as is well known, differs from most spices in that water and weak alcohol extract a larger amount of solid matter from it than does 95 per cent alcohol; but these additional extractives add nothing to the value of the extract since the aromatic principles of the root, the essential oil and oleo resins are almost insoluble in dilute aleohol and water. Street! has called attention to the importance of determining the alco- hol-soluble solids of ginger extracts as a means of judging of their purity and strength, since he finds that practically al] of the solids in high grade extracts are soluble in 95 per cent alcohol. No reference is made to the fact that this is also true regarding the solubility in ether and since some of the adulterants of ginger extracts, notably sugar, molasses, glycerin, caramel, ete., which would be more or less soluble in alcohol, would be practically insoluble in ether, it was decided to study the solubility of the solids in ether on extracts of known composition, made with weak and strong alcohol, and to compare the results obtained with the alcohol- soluble solids. It was believed that owing to the insolubility of the named adulterants of ginger extract in ether, this determination would show adulteration and low strength even better than the alcohol-soluble solids. It was our purpose also to show from the results, the relative strength of an extract. Eleven samples of ginger root, representing three different varieties, were accordingly procured, ground, and analyzed with the following results: TABLE 1. Analysis of ground ginger. ASH ETHER EXTRACT NO. VARIETY ALC OHOL Total | SOMGie | indldbie| Total | Volatile | oti. | per cent per cent per cent per cent per cent per cent per cent A PAtTCan! ayer cry 5.85 3.40 2.45 8.15 1.50 6.65 5.42 AEG Olrucroese ac oe 4.40 2.65 1.75 7.60 1257 6.03 5.88 3 ed Obs Pits orsaaicee 5.40 Saad 1.83 Calle 159) 5.62 5.93 A@ochinke-cer acc 5.08 2.80 2.28 6.15 125 4.90 lyaale¢ 5 Onbhiacteewesstacot 4.88 2.88 2.00 8.65 2.08 6.57 7.00 (| Clot, GUAR ae 5.00 2.85 2s 6.00 il il? 4.83 6.00 (AL, CROs legen seein 4.58 2.88 ITEC O| | ceo Sooner (crs ue Fo 4.98 §.13 ll UeoeniGhy soe aaa ne 3.62 2.82 OPSOM |S erotic cee ee 3.08 4.29 9} Jamaica, bleached| 6.30 2.82 SPAS Ce Sek ine te le terey ree 3.63 4.36 10|\ Jamaican. .: 2:5...) 4.10 2.85 MDB to] Ere tuavaea tote llwcaatateeea 4.70 4.40 TET] Ro Kae Seve ae 5.75 4.02 VASA Ae Gis, toc otto eriess 3.12 5.33 Maximum........| 6.30 4.02 3.48 8.65 2.08 6.65 7.00 Minimum........| 3.62 2.65 0.80 6.00 1 Le 3.08 4.29 Average.......... 4.99 3.05 1.95 7.45 1.52 4.74 5.36 1 Bur. Chem. Bul. 1387, p. 76. 508 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 8 Two sets of extracts were then prepared from these, according to the method of the U. S. Pharmacopeia using 95 and 50 per cent alcohol; these were analyzed for alcohol, total solids, aleohol-soluble solids, ether-soluble solids and the ratio of the alcohol-soluble and ether-soluble solids to total solids calculated. The following results were obtained. TABLE 2. Analysis of ginger extract made with 95 per cent and 50 per cent alcohol. SOLIDS (GRAMS PER 100 cc.) RATIO RATIO ALCOHOL ALCOHOL- ETHER- xo. vannery a ea EN PERERA eC - T- VOLUME) Total SOIR bIS solubl B TorOrey pve 95 PER CENT ALCOHOL: 1 Africantkreer sen cere ne 92.43 | 1.28 1115S 1-20) | e075 | leo 2 oO 92.43 1.22 1.18 ilealy/ 112203") L104 3 do 92.75 1.14 1.06 1.06 1:1.08 } 1:1.08 4 Cochin.... 91.36 1.04 1.01 0.99 12003) | eE05 5 ) 91.78 | 1.50 1.40 135) |) 11207 | Sie 6 do 92.74 | 1.04 1.03 O297 |e O1) | elo 7 do 92.00} 1.16 1.08 D505) |) 12107 | ero 8 JETTED sae oo ap opoceds 92.43 | 0.80 0.79 0.71 POs | Pus 9 Jamaica bleached..... 93.12 | 0.93 0.87 0282) || Ls O7 Raat 10 TAINS Aves eons oe 93.12 | 0.85 0.80 0.76 L106) | tse 11 do 93.44] 1.14 1.04 1.00) | 1:1.09) |) Lia Maximus peace eee 93.44 1.50 1.40 1.35 1:1.01 | 1:1.02 Minimium’:.seesereecieeee 91.36 | 0.80 0.79 0.71 1:1 09) | ae Average........- match ate 92.51 | 1.09 1.04 1501] 0517055) eos 50 PER CENT ALCOHOL: 1 Airicansee as: seme 49.60 | 1.90 0.79 0.382 | 1:2.41 | 1:5.94 2 do 49.44] 1.82 0.76 0.381 | 1:2.39 | 1:5.87 3 do 49.01 2.15 0.87 0.27 12°47 | 1e7e 97 4 (Coys niagsesousocscnen le Dacitl ler (a) 0.98 0.38 | 1:1.72 | 1:4.48 5 do 49.60 | 2.17 1.07 0.54 | 1:2.03 | 1:4.02 6 do 49.60 | 1.59 1.05 0.40 | 1:1.51 } 1:3.98 7 do 49.44] 2.18 0.78 0.38 | 1:2.80 | 1:5.74 8 AaMaAcia-- eee eee saees 49.17 1.93 0.34 0.19 1:5.68 }1:10.16 9 Jamaica bleached..... 48.00 | 2.15 0.35 0.29 1:6.01 | 1:7.41 10 VamMaicasssceeecen eee Al Via aed 0.31 0.24 | 1:7.00 | 1:9.04 11 do 49.33 | 2.47 0.59 0.37 | 1:4.18 | 1:6.68 Misr yeaa 49.71 2.47 1.07 0.54 Letra esses MbiibaniNesopgsccooseecos 47.17 1.59 0.31 0.19 1:7.00 |1:10.16 AV erarels fis jeuvocncras sere 49.09 | 2.02 0.72 0.34 | 1:3.47 | 1:6.48 A study of these analyses shows that in the strong extract the total solids vary from 0.8 to 1.50 grams per 100 ec., the alcohol-soluble solids from 0.79 to 1.40 grams, the ether-soluble from 0.71 to 1.35 grams, the ratio of aleohol-soluble to total solids from 1:1.01 to 1:1.09, and the ether- soluble to total solids from 1:1.02 to 1:1.14. It appears, therefore, from these results that in an extract prepared with 95 per cent alcohol, practically all of the solids are soluble in ether as well as in alcohol, and the ratio of | 1916] HARRISON AND SULLIVAN: STANDARD GINGER EXTRACT 509 the ether-soluble solids to the total solids is nearly the same as the ratio of the alcohol-soluble solids to total solids, being a trifle higher. The results on the extract prepared with 50 per cent alcohol, however, are quite different. As would be expected, the total solids are consider- ably higher and the solids soluble in alcohol are as a rule about 40 per cent of the total solids, with an existing ratio ranging from 1:1.51 to 1:7.00, with a general average of about 1.2:5. In the case of the ether-soluble solids, however, a much higher ratio exists since the ether-soluble solids are much lower than thealcohol-soluble solids (generally about 50 per cent) and the ratio of ether-soluble solids to total solids is found to range from 1:3.98 to 1:10.16, being generally above 1:5. It therefore appears that while the solids from a high grade extract are almost completely soluble in ether, only a very small amount of the solids from an extract prepared with weak alcoho] are ether-soluble, and that, with a very dilute extract as the analysis before quoted where the alcoholic strength was 24 per cent, there were no ether-soluble solids, and such extracts are therefore practi- cally valueless. Ginger extracts prepared with 50 per cent alcohol have, as compared with extracts made with 95 per cent alcohol, about twice the amount of solids, 70 per cent of the amount of alcohol-soluble solids and 33 per cent of the amount of ether-soluble solids. SOLIDS SOLUBLE IN GINGER, NO. 6 Alcohol Alcobol Alcohol Alcohol Water 95 per cent 75 per cent 50 per cent 25 per cent eS per cent per cent per cent per cent per cent Solids soluble in 6.00 8.92 9.20 9.82 13.39 ether 6.00 4.10 2.76 0.10 0.1 It is self-evident from these tabulated figures that strong alcoho] must be used to get the full flavoring strength of the ginger. The official methods were followed in the determination of alcohol, solids, and alcohol-soluble solids as given in the 1910 Proceedings (Bur. Chem. Bul. 137, p. 79). For the ether-soluble solids 10 cc. were evapo- rated in a porcelain dish to complete dryness. Absolute ether was then added to the residue, the dish covered with a watch glass and allowed to stand 15 minutes. This ether was then decanted through a dry filter into a tared 100 cc. Erlenmeyer flask, and the ether washing repeated. The undissolved solids remaining in the dish were then scraped loose from the sides with a spatula and rubbed up with successive small portions of ether which were passed through the filter until no more material was dissolved, as indicated by the ether coming through colorless. The ether was then distilled off and the flask dried at 100°C. to constant weight. 510 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 3 SUMMARY. From the analytical results the following conclusions are drawn: The solids in ginger extracts made with 95 per cent alcoho! are practically all soluble in ether, this being true of all three varieties examined. Only a small amount of the solids from ginger extract made from weaker alcohol is soluble in ether, and the solubility of the solids in ether decreases the weaker the alcoholic extract, even though the amount of total solids increases, and in the case of the 25 per cent extract, practically none of the solids were soluble in ether. The solids from extracts made with weak alcohol are more insoluble in ether than in alcohol. Some of the adulterants of ginger extracts, as molasses, caramel, glycerin, sugar, ete., would be more insoluble in ether than in alcohol, and, therefore, the percentage of solids soluble in ether conveys more information as to the strength of a commercial ginger extract than the amount of alcohol-soluble solids. The amount of ether-soluble solids in connection with alcohol-soluble solids is a valuable measure of the strength of the extract. REPORT ON SPICES. By R. W. Hits (Bureau of Chemistry Food and Drug Inspection Laboratory, San Francisco, Cal.), Associate Referee. At the 1913 meeting of the association, there were referred to the associ- ate referee on spices, the recommendations of the associate referee on condiments other than spices for 1911, as the refereeship on condiments other than spices was dropped in 1912. These recommendations con- cerned in particular the examination of tomato ketchup with special reference to the detection of spoilage, and the work was continued in 1912 when the methods for the determination of lactic acid, citric acid, insoluble solids, and sand were approved for final action as provisional in the following year. The present associate referee has had no oppor- tunity to make a special study of the methods, but would state that the methods for citric acid, Jactic acid, and insoluble solids (Bur. Chem. Bul. 162, pp. 128-129) are used in the San Francisco laboratory of the Bureau of Chemistry, as occasion arises, for the examination of ketchup. While the lactic and citric acid methods appear to have a value in detecting the use of decomposed pulp, this value is only as confirmatory of the micro- scopical methods, which must still be the main reliance. The lactic acid method is laborious, and, in a mixture such as ketchup, must be re- garded as merely conventional, and its results must be taken as relative and not as absolute. —— 19165] REMINGTON: BAKING POWDER 511 With the idea of finding some method that might yield lower ‘blank”’ results on ketchup, a few preliminary experiments were made to test the availability of the method of Méslinger,! which depends upon the solu- bility of barium lactate in 70 per cent alcohol. At present it appears probable that the complications caused by the presence of sodium benzoate might make the method quite cumbersome, even if applicable. Certain facts deserve consideration: The lactic acid method which has been under study is Jaborious and its results, as well as those of the citric acid method, are relative. Research work on the determination of lactic and other organic acids is under way in the Bureau of Chemistry. The lactic and citric acid methods have received the approval of the Secre- tary of Agriculture for official use, so that immediate adoption is not necessary. It is recommended that final action regarding the lactic and citric acid methods be withheld pending further study and collection of data and that the methods for determining insoluble solids and sand, as applied to ketchup (Bur. Chem. Bul. 162, pp. 128-129), be adopted as provisional. REPORT ON BAKING POWDER. By Ror E. Remineron (Agricultural Experiment Station, Agricultural College, N. D.), Associate Referee. Early in the year the associate referee sent out about fifty copies of a circular letter suggesting four determinations for study on baking powders. From replies received, it seemed that the question of lead was of most interest, and after some correspondence with the associate referee on heavy metals it was decided to take up the study of this determination in baking powders, particularly those of the alum-phosphate type. Correspondence with a number of chemists and a study of Jast year’s report on Jead, Jed to the conclusion that our present methods are not sufficiently perfected for collaborative work on samples. Accordingly, this year’s study has taken the direction of attempts to simplify and improve existing methods to the point where they may be placed in the hands of a number of chemists for check determinations. The associate referee has had the codperation of several chemists in this study, and begs to submit the following modification of existing methods as a result of the year’s work. He would recommend for next year that a comparative study be made of the Seeker, Exner and referee’s methods for this determination. The latter is as follows: 1 Zts. Nahr. Genussm., 1901, 4: 1120; 1914, 27: S41. 512. ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 3 REFEREE’S MODIFICATION OF SEEKER METHOD. Weigh 100 grams of baking powder into an 800 ec. Jena beaker and add 75 ec. of hydrochloric acid in small portions, with stirring, followed by 200 ce. of distilled water. Place the mixture on a steam bath and heat until the starch is completely hydrolyzed and the solution is fairly clear and a bright yellow. (Too long heating will result in darkening due to caramelization. A copious crystalline precipitate of calcium phosphate may appear, but will do no harm.) Neutralize the solution to the point of precipitation of aluminum with ammonia, cool, and add 4C0 cc. of a 20 per cent ammonium citrate solution saturated with hydro gen sulphid. This citrate solution is prepared as follows: Prepare a solution containing 200 grams of ammonium citrate per liter in a bottle fitted with a two-hole stopper. Through one hole, make connection with a generator producing hydrogen sulphid, and to the other attach a siphon for drawing off the solution. The liquid is thus kept saturated with hydrogen sulphid, and at the same time may be drawn off with- out disturbing the precipitate of lead sulphid which has gathered on the bottom. Practically all the ammonium citrate on the market contains lead. By preparing the solution as just described the loss due to solubility of lead sulphid is reduced to a minimum, the reagent being already saturated with this salt. The baking powder solution will be slightly alkaline, so that a partial precipita- tion of iron sulphid will result, together with that of the lead. Allow the solution to stand covered, passing in more hydrogen sulphid if the odor of the gas disappears, until the precipitated sulphids settle. Siphon or filter off the supernatant liquid, collect the precipitate on a filter and wash with a small amount of hydrogen sul- phid water. Then place the filter with the precipitate in a tall 100 ec. Jena beaker, add 5 ce. of sulphuric acid and 10 ee. of nitric acid, and heat on the hot plate with occasional additions of nitric acid until the solution is colorless and all organic mat- ter is destroyed. The solution will contain a precipitate of calcium sulphate. Con- tinue the heating until a large portion of the sulphuric acid has passed off, but not to dryness. Cool, add 5 ec. of water and heat gently to aid solution. Again cool, neutralize with ammonia, and add 30 ec. of a mixture of equal parts of saturated lead-free ammonium acetate solution and 96 per cent alcohol. Allow to stand over- night, filter and wash with a small amount of ammonium acetate alcohol solution. If the precipitate of calcium and ammonium sulphates is heavy, it may be necessary to use suction. Acidify the filtrate with acetic acid and heat nearly toboiling. Precipitate with potassium dichromate solution, allow to stand overnight, filter on a small tared Gooch, wash with a small amount of cold water, dry at 125°C. and weigh as lead chromate. No report was made by the associate referee on meat and fish. 19165] KERR: FATS AND OILS 513 REPORT ON FATS AND OILS. By R. H. Kerr (Bureau of Animal Industry, Washington, D. C.), Associate Referee. The 1914 work consisted of a study of methods for the detection of phytostero] in mixtures of animal and vegetable fats. Two methods were studied, the one developed in the laboratories of the Bureau of Animal Industry and the digitonin method of Marcusson and Schilling. Three samples were sent out for the work, the composition of them being as follows: Sample 1—Lard adulterated with 5 per cent of cottonseed oil and 0.25 per cent of vaseline. Sample 2—Pure lard (rancid). Sample 3—Lard adulterated with 2.5 per cent of hydrogenated cotton- seed oil and 2.5 per cent of soy bean oil. Vaseline was added to Sample 1 in order that the effect of its presence on the methods studied might be determined. Such an amount of vaseline would effectually prevent the obtaining of any accurate results by the present provisional method. A rancid lard was chosen for Sample 2, for simiJar reasons. Rancidity interferes decidedly with the present pro- visional method. The following instructions were sent to the collaborators: INSTRUCTIONS TO COLLABORATORS. The work consists of a study of methods for the phytosterol acetate test. Three samples of fats for this work are being sent you under separate cover. You are re- quested to test these samples for phytosterol according to the methods given below. 1. Bureau of Animal Industry method.—Described in Bureau of Animal Industry Circular 212, a copy of which is being sent you herewith. 2. Digitonin method of Marcusson and Schilling—(Chemiker Zeitung, August 21, 1913.) Shake vigorously for 15 minutes in a separatory funnel 50 grams of the fat oil or fat to be tested with 20 cc. of a 1 per cent solution of digitonin in 95 per cent alcohol. Allow the mixture to stand for a time until the emulsion separates. The lower or fat layer should be quite clear while the alcohol layer is full of a bulky, flocculent precipitate. Draw off the fat as much as possible, taking care not to lose any of the precipitate. Add 100 cc. of ether to the alcohol layer and filter the mixture. Wash the precipitate with ether until free from fat; after drying in the air, transfer it to a tall 50 ce. beaker, add 2 to 3 ce. of acetic anhydrid and cover the beaker with a watch glass. Then boil slowly over a low flame for half an hour. After cooling, add 30 to 35 cc. of 60 per cent alcohol and thoroughly mix the contents of the beaker. Filter off the alcohol solution and wash the precipitate with 60 per cent alcohol, then dissolve it on the filter with a stream of hot 80 per cent alcohol from a wash bottle and set away the filtrate in a cool place (10° C. or below). After the acetates have crystallized out of this solution filter them off, recrystallize from absolute alcohol, dry, and determine the melting point as divected in Bureau of Animal In- dustry Circular 212. 514 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 3 You are requested to test the three samples sent you for phytosterol by both methods and to report (1) melting point of acetates obtained in first crystallization from absolute alcohol, (2) melting point of each subsequent crop of crystals, if ace- tates are recrystallized, (3) judgment as to presence of vegetable oil, and (4) which of the two methods you consider preferable. Any comment or suggestion you may care to make will be appreciated. RESULTS OF COLLABORATORS. Results by L. B. Burnett (Bureau of Chemistry). Melting points by B. A. I. Melting points by Digitonin Sample method method Conclusion 1 114.8°—Ist crop crystals 114.8°—I1st crop crystals Phytosterol present 115.6°—2d crop crystals 116.0°—l1st crop crystals 115.8°—1st crop crystals 2 111.8°—1st crop crystals 111.4°—1st crop crystals No phytosterol 112.0°—2d crop crystals 112.2°—1st crop crystals present 3 115.8°—Ist crop crystals 116.0°—1st crop crystals Phytosterol present 116.4°—2d crop crystals 116.6°—1st crop crystals Mr. Burnett expresses a preference for the digitonin method over the B. A. I. method and recommends using a 100 gram sample for the test. Results by R. S. Hollingshead (Bureau of Chemistry). Melting points by Digitonin method 116.5°—1st crop crystals 116.1°—2d crop crystals Conclusion Sample 1 Vegetable oil present 2 114.2°—1st crop crystals Pure lard 113.6°—2d crop crystals 3 115.8°—1st crop crystals 115.4°—2d crop crystals Mr. Hollingshead reports no results by the B. A. I. method. preference for the digitonin method on the ground of convenience. Vegetable oil present He expresses a Results by R. H. Kerr (Bureau of Animal Industry). Melting points by Digitonin method Conclusion 117.4°—1st crop crystals Vegetable oil present Melting points by B. A. 1. method 115.4°—I1st crop crystals 116.2°—2d crop crystals Sample 2 113.6°—1st crop crystals No vegetable oil 114.0°—2d crop crystals 114.4°—1st crop crystals 3 117.0°—I1st crop crystals 117.6°—l1st crop crystals Vegetable oil present These results show no choice between the two methods with regard to accuracy. Each method led to uniformly correct conclusions. Choice between the two methods must then depend on other factors. The digitonin method offers the advantage of simplicity and convenience and has the disadvantage of demanding an expensive reagent, which also is limited in supply, and obtainable only with difficulty. The B. A. I. method lacks this disadvantage, but requires more time and labor for its manipulation. Both methods are decidedly superior to the present provisional method. 1915] LADD: PRESIDENT’S ADDRESS 51 RECOMMENDATIONS. It is recommended— (1) That the B. A. I. method for the detection of phytosterol in fats (B. A. I. Cir. 212), be adopted as a provisional method. (2) That the digitonin method as described in this report be also adopted as a provisional method. (3) That the glycerin saponification method for the preparation of fatty acids for use in the titer test (Bur. Chem. Cir. 108, p. 11) adopted last year as a provisional method, be made official. (4) That Emery’s method for the detection of beef fat and other solid fats in lard (B. A. I. Cir. 132) adopted last year as a provisional method, be made official. PRESIDENT’S ADDRESS. By E. F. Lapp (Agricultural Experiment Station, Agricultural College, N. D.). MEMBERS OF THE ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS: For the thirty-first time we, as members of this association, are gathered together to carry forward the work started by that little group of chemists who organized the association nearly a third of a century ago. Even though possessed of prophetic vision, little could they have realized the great benefits that would grow out from their modest beginnings, and the vast amount that would be accomplished in promoting agricultural chemistry in this country, or, how far-reaching would be the influence of the work of this association. The methods that have been developed and approved during these years have not only come to be generally recognized as official in American work and so accepted by the courts, but they are generally known and used in European countries. And yet, there remains before the association a great field, rich for the harvest, and chemists are anxiously awaiting the information. The regulatory work demanding the attention of the chemists coming through the enactment of new laws, has made a demand for more exact analytical methods in many fields of activity. Where we had formerly fertilizer laws, then feeding stuffs laws, we now have food laws, drug laws, insecticide and fungicide laws, sanitary laws, and moving rapidly towards us, is the perplexing work with paints, varnishes, textile fabrics and scores of other products that the official chemist must be prepared to handle. In many of these fields we shall need to caJ] for aid upon the physical chemist, the biological chemist, the microchemist, the toxicologist, the physiologist, the bacteriologist, in fact, upon the whole realm of science to bring to bear their skill and knowledge in solving the problems of the people and to hold the prestige that is ours as workers in this great field now spread out before us. 516 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 3 The splendid work which has been done in the past and the commanding position for our association and its workers has been made possible by the generous aid of the Bureau of Chemistry in all of our undertakings, and, to a great degree, by the publication by the Secretary of Agriculture, of our past proceedings as well as of the methods of analysis, as now given in Bulletin 107, Revised. Our work can never end for each year presents us new problems to be solved, new investigations to be undertaken, new lines of research to be developed, which call for new and improved methods of analysis in the great field of agriculture as well as in the official regula- tory work for protecting the public health and in the prevention of fraud and deception. The work of this association has gained a place because the information resulting from these gatherings has been placed in an available form before the public for the use of our chemists. It is most unfortunate, therefore, that the Department of Agriculture has, by legal restrictions, been unable to continue the publishing of these important proceedings and, especially so, since the methods approved by the associ- ation, are now generally recognized as official under both State and Federal regulatory laws and, at a time, when we, as official chemists, need all the chemica] information possible to be secured to aid us in preparing for our official duties. DISCUSSION OF PUBLICATION. As you are aware, the proceedings for the past year have not been published for lack of funds, and, since satisfactory arrangements therefore have not been perfected with any of the large publishing houses, it is vitally important that some means be found for a prompt publishing of the work of the association if we would continue to make progress, as is most essential at this time. A series of questions were sent out to a number of chemists to get their viewpoint, and the three questions propounded on this point were: 1. How shall the proceedings of our association be published and made available for the general use of chemists? 2. Should the association attempt to publish a journal to include the proceedings and methods of analysis? 3. If an association journal be published should it attempt to care for other than official matters of the association, or should it include articles of interest to mem- bers of the association, including editorial matter of general interest? Fifty letters were sent out, to which there have been twenty-five replies. I can only in general terms analyze the statements made by the several writers without giving due credit to the authors in each instance. Five only were strongly in favor of publishing a journal. Twelve believed the association should publish an annual report of its proceedings. It is thus observed that but few are favorable to publishing a journal, chiefly on 19165] LADD: PRESIDENT’S ADDRESS 517 the ground of expense, and, for the further reason, that there are already a large number of chemical journals available. It should be borne in mind, however, that there is no definite publication devoting its attention to the particular field of work which this association represents. Several of the chemists were strongly in favor of again urging the Department of Agriculture to publish the proceedings, as in the past, since the work is largely of an official character and must be the basis for analytical work in connection with the enforcement of food and drug laws, both State and Federal. On the other hand, there are those who are equally strong in favor of the publication of a journal quarterly to include the proceedings of the association and admit that there are many financial difficulties in the establishment and maintenance of a new journal, but believe that the labors of the agricultural chemists will justify the establishment of such a journal. Mr. Frear stated: The Association of Official Agricultural Chemists is probably the most broadly representative organization comprising the interests of agricultural chemists. Considering the volume of its present labors, I should not think it wise to enlarge the scope or program of its meeting, but I do not see why that should limit the scope of its journal to purely official subjects. As a matter of strategy, I would suggest that the proceedings of the association appear in sections, separately paged, and added as a supplement to the journal, and that the proceedings be distributed through the numbers of the journal in such a way as to promote its circulation among the various laboratories of the country, whose chemists need to keep acquainted with the changes in official methods and the reasons therefor. A separate pagination in this way would permit of its binding into a separate volume. Mr. Fraps stated: The proceedings should be published either as a single publication or in periodical issues. It does not appear to me that the subject matter lends itself readily to periodical treatment, but would better appear as a single publication. Mr. Alsberg, who has been acting for the committee appointed at the last association meeting, to examine into the matter of publishing the proceedings, stated: On the whole it seems to me that the reports have not been very encouraging. Personally, I feel that the journal could serve a valuable purpose in two ways: First—by printing a lot of material in the way of data and analyses which are un- available to food, drug and feed control officials, and which are more in the nature of statistics and, therefore, not available for publication in ordinary research jour- nals; second—by the stimulus it will give to scientifie research. Analyzing the data as gathered thus far by Mr. Alsberg, it would appear that there are 98 subscriptions promised and 20 probable subscriptions, 518 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 3 or, a total of 118 subscriptions. This is hardly a sufficient guarantee to warrant one in proceeding with the publication of the journal. The matter is, therefore, at this time placed squarely before the association as to what steps shall be taken with regard to the publication of the pro- ceedings of this association. Shall the association attempt the publication of its proceedings in pamphlet form, to be furnished to the members of the association, and sold to those interested in the work, or, shall the association take steps looking toward the publication of a journal or - periodical, say quarterly, and thus be able to avail itself of the postal rates, at the same time furnishing a medium for the publication of a vast amount of information of great importance to the agricultural and food chemists, and in this way stimulating research work in agricultural and food chemistry? For example, like the Analyst of London for the food work and supplements each quarter to contain the proceedings and revised methods of analysis, or, after the fashion of the Zeztschrift fiir Analytische Chemie. In either case, how can we finance this undertaking so as to ensure success? As an association we can not afford to neglect publishing our proceed- ings, but it, does seem to me that we may well edit out a portion of the material in the published proceedings and summarize more fully the data. Many think foo much time is given to hearing the reports of the referees, and not enough to the discussion of results, or to the methods which are proposed for the use of the association. Professor Christensen of New Mexico thinks it would be far better, instead of publishing a new journal, to try to affiliate with the American Chemical Society in such a way that some one of the journals of that society would create a section on agricultural chemistry to be devoted to articles bearing on the various phases of agricultural chemistry and such other matters as may be of special interest to the Association of Official Agri- cultural Chemists. Undoubtedly, the secretary will be able to present to you some further information with regard to the possibilities of publishing the proceedings of this association. REFEREE AND COLLABORATIVE WORK. With the vast number of research problems confronting the members of this association, the question is raised as to how we shall secure the best possible results in a reasonable length of time. There are those who have thought that some improvement might be made in handling the vast amount of work now in the hands of the referees. Some have felt that too much time, at the association meetings, is taken up in the presentation of formal reports and the immense amount of detailed analytical data quite impossible for the hearer to follow; that not enough time is given to 1915) LADD: PRESIDENT’S ADDRESS 519 the discussion of important matters before the association. If this be true, how we shall remedy it is a perplexing question, and it is my purpose to point out some of the opinions expressed by our members in order that the matter may be brought more directly to your attention; trusting, in this way, that we may be able to find a remedy. None seem to think it desirable to discontinue our present methods, especially, the sending out of samples to try out methods, but the chief point seems to be that too large a field of work is assigned and that by restricting the work to a narrower field or to a single problem, more could be accomplished. Again, not enough interest is taken in many of the laboratories, my own included, to accomplish what should be done in collaborative work. The least experienced, in reality, apprentices, too often are entrusted with the analytical work of trying out the methods because our laboratories are overcrowded and so render it practically impossible for the better trained men to do work of this kind. So the work is slow and, at times, possibly methods have been condemned be- cause there has not been a fair try-out by well trained chemists. Can we not accomplish more by having two types of work? Let chemists be appointed individually, or in group committees of three, as may seem most desirable, chemists who are known to be interested in certain fields, to take up the necessary research questions and to develop methods; and, when this much has been accomplished, turn the matter over to the referees that the methods may be tried out by a few select workers in the same field before the testing-out process begins by sending out samples. By this method it seems to me that a half-dozen research problems could be attacked by different workers, where we now have only one. Methods could be developed which would be far more satisfactory and the work done much more rapidly than by the present system. Let us not, however, lose any of the essential benefits arising from the splendid line we have already developed. As illustrating the views of different chemists, I quote from their statements. One stated: The only criticism I can make of the work of the referees is that they have in many cases been assigned a great subject which involved a lifetime work, or, at any rate, work for many years to cover it completely. It would be a big improvement in my opinion if, instead of assigning general subjects, little portions of that sub- ject be assigned by the Executive Committee, according to a definite plan, so that only one small investigation at a time is occupying the referee. Then thewhole subject will be gradually covered by the referee, or a series of referees, in the course of a few years. ‘A subject like feed adulteration, for instance, in general terms, without any spe- cific problem involved, does not seem a profitable way to get any clean-cut results. More, I think, could be accomplished by selecting one or two definite forms of adul- teration and instructing the referee to work and report on that instead of letting him wrestle with the whole subject. 520 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 3 I feel very strongly that the referee should be given the greatest amount of free- dom in the handling of his subject. I also feel, however, that in outlining his work he should only plan for as much as he feels he will be able to complete during his two years asreferee. Then, if it is the feeling of the association that he has rendered very valuable service, he might be reappointed to continue the work, I think, however, that this latter course is only advisable in unusual cases. The question of sending out samples is largely dependent, it seems to me, upon the work in hand. In many eases it will doubtless be necessary for the referee to devote his entire time to research. Again, he may have his methods worked out after a few months to such a point that he feels the need of codperation and confirmation of his own results. In such a ease I think too much emphasis cannot be laid upon sending out samples. I do feel, however, that the samples should not be sent out until after the referee has done a great deal of research and worked on the method fora sufficient time to have a very definite opinion in his own mind as to the practicability of the method. I feel that this is necessary owing to the fact that the referee, after working with a method for a long time, often becomes so familiar with it that in his hands it gives very satisfactory results, whereas, in another’s hands, it may not prove satisfactory. Another, connected with one of the leading experiment stations stated: It has never seemed to me that the present method of trying out analytical meth- ods has amounted to much. Progress in analytical chemistry has come from the specific and individual interest of a chemist in a given research and for which it has been necessary to develop the best methods possible. The present method fails to bring out the critical attitude of the analyst, but merely blindly follows printed directions. If assignments could be made to those individuals particularly interested in the method under investigation, then, I am sure, very much would be accom- plished. Another stated: In the development of new methods investigation is necessary, but in the testing of methods proposed for adoption, the collective testing of methods is necessary. The system of sending out samples is to a certain extent overdone, as sometimes methods are sent out which have not been tested sufficiently by the referee himself. Another station chemist stated: It is our belief that assignments by a referee of work to a laboratory which has become known because of its special attention to that subject, would lead to better results than by the present system. It is our belief that in a great many cases poor results are secured by the use of a good method because of the inexperience of the worker. Another Western chemist stated: I think the assignment of subjects for investigation to individuals might be pro- ductive of a higher grade of work, also new and better methods. 1915] LADD: PRESIDENT’S ADDRESS 521 An Eastern chemist stated: I like the present system of sending out samples by the referees to try out methods, but think that the methods should be thoroughly tried out by the referee before sending out his samples. The trouble with this method seems to be that much of the work is now being done by the younger and more inexperienced chemists, making the matter rather a trial of men than methods. Another: For improvement in the work of the association, my suggestion would be to select some subjects and assign them either to committees or to individuals. We have already made a beginning in this direction in the matter of methods for basic slag. I feel that we should make further progress along this line, at the same time, continue to study methods thoroughly by sending out samples before undertaking to adopt the methods proposed. i Another: It has seemed to me that too much time was taken up with reports on analytical methods with its tiresome array of figures, and not enough upon some new phases of agricultural chemistry, possibly even outside of analytical work. In other words, not enough time is set aside for the presentation and discussion of papers dealing with new investigations in our field. The general tenor of the statements from the other chemists is similar to those already presented and indicate that there is a belief that at this time we can make progress more rapidly by inducing chemists who are interested in some special field of work to take up some individual problem and develop methods satisfactory therefor; and then, possibly, these methods be turned over to the referees who will proceed to try them out after they have been carefully tried out by associate chemists interested in the same field of work. I would also call your attention to the desirability of increasing the membership of the Association of Official Agricultural Chemists. And in this connection would it not be desirable to admit to membership chemists representing municipal organizations, who work along the same line as that of our association? Iam informed that the secretary has had several requests for membership in the association by official municipal chemists, yet, according to our Constitution, these men are not entitled to admission. Should the Constitution be amended so as to admit to membership chemists who are engaged in municipal work? Other suggestions have been made as to how to improve the efficiency of the association, and among these suggestions is one that the referees take up too much time of the association in submitting their reports. Should the report of the referee be restricted, say, to 15 minutes? If this be done only an abstract of most reports would be presented to the 522 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 8 association, and in this event the Committee on Recommendations would have to go over the whole report in more detail than they do at the present time, thus, throwing more work on the committee, but it is the belief of a number of the members that such would tend to make the meetings more interesting and allow of a fuller discussion of subjects before the association. It has been suggested also that a committee be appointed to go over very carefully the official methods of analysis as now published and in- clude among the revised methods only those which have been actually demonstrated to be of value and dependable. One chemist with regard to recommendations made the statement: I think it is about time to get some method for determining available phosphoric acid in fertilizers which makes use of a more dependable reagent than so-called neutral ammonium citrate. I think it would be well to take up and push rapidly some work along this line. Perhaps, try out some of the methods proposed by the fertilizer section of the American Chemical Society. The general tenor of a number of the replies is that the methods and important data of the convention are not gotten out, as early as they should be and made available for the use of official chemists. This im- provement could best come through the publication of an official journal by the association. Ifa quarterly journal could be published under the auspices of the association, I feel confident much good would result and a greater stimulus would be given agricultural chemistry, such as has never been known in this country. Our workers would be more closely welded together and the benefits would be far-reaching. If, too, the Chemical Section of the National Association of Dairy, Food and Drug Officials could be induced to coéperate with us and use one issue for the proceed- ings of their section, it would, in my judgment, be a most desirable aid to our official work. Some method must needs be devised for publishing the proceedings if we would continue to make growth, and I strongly urge upon all our members to consider carefully at this time the best method to handle this important matter. RECOMMENDATIONS. As suggestions for the consideration of the association and for the betterment of the work in order to enable us to accomplish the greatest amount of good and to insure the fullest value in our undertakings I recommend— (1) That there be appointed a committee of four or five members who are familiar with the work of the association and the problems con- fronting them for the purpose of receiving suggestions from members of the association with regard to problems to be solved and that the said committee canvass carefully the suggestions made, formulating there- 1915] ALSBERG: PROPOSED JOURNAL OF AGRICULTURAL CHEMISTRY 523 from plans for the future work of the association, determining what par- ticular problems are best taken up and find possible chemists interested in this particular field of work and assign to them such phases of the work as may be carried out in a reasonable period of time, keeping in mind a definite policy for the development of the work. This committee might well be composed of one man fully familiar with soils and agriculture problems, another with fertilizer problems and another member to be familiar with the questions involved in food aad drug work. (2) That the association take action at this meeting looking to the publication of the proceedings and it seems to me after canvassing the situation carefully that a quarterly will afford us the best meansof placing before the chemists the work of the association and enable us to take ad- vantage at the same time of lower postal rates. (8) That a time limit of 15 minutes be placed upon the papers and reports and that members be urged to condense as fully as possible the material that is necessary to report leaving the great mass of figures to the consideration of the committee or referee. (4) That the time has come when the work might well be divided for one half day and two section meetings held, one dealing strictly with agricultural problems and the other with food and drug work. In this way the two sets of chemists would be permitted to select such meeting as they deem most nearly fitting their particular line of work; all other papers and discussions would then be presented in the full session. This would shorten the period of the meeting and permit discussion on the part of members. (5) That there should be introduced more material which deals with problems, short papers or addresses that would be of general interest to all the members and that the time has come when our work should not be confined so exclusively to detailed methods of analysis but that there should be added constructive suggestions to encourage and stimulate investigation on the part of our younger members. REPORT ON PROPOSED JOURNAL OF AGRICULTURAL CHEMISTRY. By C. L. Atsprre (Bureau of Chemistry, Washington, D.C.), Secretary. After giving much time and thought:to the matter of publishing the proceedings of the association, and consulting with many members of the association and with various printers, I have reached the conclusion that, considering the high standing of the firm, and its great experience in pub- lishing research journals, the proposition submitted by the Waverly Press of Baltimore, is a very fair one and the best we ean secure. Unfortu- nately, few publishers are ready to assume even so great a risk as this. 524 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 3 This firm promises to publish a quarterly journal to be known as the Journal of Agricultural Chemistry, to be restricted to agricultural, food, and drug chemistry, to contain not less than 600 pages per volume of 4 issues; to print, bind, publish, issue, mail and distribute same; to edit manuscript; pay all expenses; solicit subscriptions and advertisements, but not to pay anything to authors for manuscript and copy for illustra- tions; to submit all matter to the managing editor for approval before incorporation in the journal; and to submit a make-up copy to said editor before publication. This journal is to be in all respects like the Journal of Biological Chemistry, published by the Waverly Press. The association, on the other hand, must agree to the use of its name in every legitimate way in the solicitation of subscriptions and advertise- ments; to furnish a list of possible subscribers and to use all reasonable efforts to further the interests of said journal. The subscription price is to be $5.00 payable in advance, but members may subscribe at $4.00. The publishers are to collect and receive all money for subscriptions and advertising and keep a list of subscribers, their records to be accessible at all times to the association. At the end of every twelve months, an account is to be rendered to the association. They are to be repaid, if sufficient funds are available, for all disbursements and expenses, and interest on money advanced, plus profit equal to 10 per cent net profit over and above all costs of material, labor, ete. They shall also receive a profit of 10 per cent on gross receipts. The 10 per cent publisher’s profits, however, are not be be paid until such time as the income of the journal warrants. The association, in turn, is to receive an amount equal to the publisher’s profit, and any remainder shall be equally divided. The association guarantees to pay any deficit up to $1,000.00 during any period of twelve months with the understanding that all advertising is to be credited to cost of printing the journal. This agreement is to be binding for five years, with privilege of renewal at expiration of such time. In case any of you are unfamiliar with the work of these publishers, I would refer you to the Journal of the Washington Academy of Sciences, the Journal of Biological Chemistry, the Journal of Pharmacology and Experimental Therapeutics, and the Journal of Phytopathology, published by this firm, all of which, as you doubtless know, have given splendid satisfaction. The publishers feel that the publication would be more attractive if it appeared in journal form, and more so, if it included the official methods which in the past have been issued in bulletin form from time to time by the Department of Agriculture. This has the advantage to the chemist of getting such official methods before the public as soon as they are adopted, instead of waiting for periodical revisions of Department publications. 1915| ALSBERG: PROPOSED JOURNAL OF AGRICULTURAL CHEMISTRY 525 The data contained in Bulletin 107, Revised, of the Bureau of Chemistry could be published either as a part of the journal or as a supplement, and the recommendations and reports of referees could be issued as they are received. In this connection, of course, the question as to whether or not the Secretary of Agriculture would be willing to relinquish the publication of official methods would have to be presented to him for final settlement. I feel that the question of getting advertising matter should be very seriously considered. If it is deemed advisable to secure advertisements, they would, of course, have to come under the strict censorship of the committee in charge of such journal or the managing editor. The pub- lishers pointed out very justly, it seems to me, that as an advertising medium for dealers in apparatus and glassware, publishers of books, manufacturers of chemicals, various types of farm implements, apparatus and machinery, this journal would have considerable advertising value. Inasmuch as all proceeds from such advertisements are to be credited to the journal printing account, the number of subscriptions necessary would be correspondingly decreased, and this is a matter well worth careful thought. It is estimated that $3,000.00 a year will be required to finance the journal, which would mean 600 subscriptions at $5.00 each, or 750 sub- scriptions at $4.00 each, provided it is thought best not to include adver- tising matter. It is also necessary that a fund be raised as a guaranty, as the subscriptions during the first year or two will probably not be sufficient to defray the entire cost of the publication. The responses received to the circular letter sent out under date of June 15, 1914, asking for subscriptions to the journal, and contributions to the guaranty fund, have not been as promising as was anticipated. A number of members of the association have written stating that they feel there are already too many scientific journals, and that it would be inadvisable to publish a journal for this association; also that the sub- scription price is a little excessive. Still others have suggested that the journal be published by the American Chemical Society, and the official methods be published by the Department of Agriculture, as has been the custom in the past. Another suggestion is that in place of calling the publication the “Journal of Agricultural Chemistry,” it be known as the “Journal of Official Chemists,” thus giving it an official status. It has also been suggested that in place of having the annual dues $2.00, they be increased to $10.00 which would include subscription to the journal. In this way the institution would pay for the dues as well as the journal, and it would not fall so heavily upon the individuals, as might be the case if the dues remain as they are at present and a separate subscription price were assessed for the journal. Still others feel that the dues should be entirely dispensed with or materially decreased. 526 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 3 The following is a brief statement of the financial status of the journal at the present time. SHDSCLIPLIONS PrOMISed hon arse elses Vale ere ete eat 98 Probable subscriptions promised...................2.--e0eee- 20 Checks in payment of subscriptions, total..................... $60.00 Cash on hand toward guaranty fund........................... 114.00 Amount promised toward guaranty fund...............-...... 205 .00 Totali¢ashion hand pee eae ae eee Meas e Ae aideac $174.00 In addition to this there have been a number of letters from members stating that they will give financial assistance in the way of launching the journal, and others have written that they will do their utmost to secure additional subscriptions. Personally, I feel that when once the journal appears, the demand for it will be very much greater than the figures just read would indicate. Iam of the opinion that we would have no difficulty in finding ten men who would put up one hundred dollars a year for five years, if need be. A detailed statement has been prepared giving the names of the individuals or organizations who have promised to contribute to this project, and should any one desire to consult this list it will be found at the registration desk. It is my personal opinion as secretary that such a journal would probably very soon be self-supporting, because it would contain very much material which would be necessary to professional and commer- cial chemists, as well as to agricultural, food, and drug chemists, and feed control officials, and I personally am under the impression that five or six hundred subscriptions could be easily secured even in the event that we do not include advertising matter. The discussion of the proposed Journal of Agricultural Chemistry was postponed until the afternoon session and the meeting adjourned at 12.40 to reconvene at 2 p.m. INDEX TO VOLUME I, NUMBER 3 PM caIMpIN SOUS PAPEL DVL ATs cerayciatnr~ civ, ©le ticle etoile )< ine ekete terete eevee aera oer Soils recommendation| by) Hare...-c.42094- 4. 2is sade eee ee PEDOLEAD Vala ne se vececite cous iistsrstekers arasis act Sie ae cnet Alsberg, note on study of vegetable proteins....................s sees eee ee ees report on proposed Journal of Agricultural Chemistry................ PRN EAMELEDOLL ON SOUS perspec es eles cei anaic ceyetelars jsie a wee sieievearelore Bm Atee carte: Ammonium citrate, neutral, report by Walker.......................ee0eee eee Auditing, committee, appointment and personnel............................. Baking powder, report by Remington....... ce Ny Seek a eae APA teal OE: oe Basic slags, phosphoric acid content, report by ogmmtnies (Ww alliemae) PEE aoe report by Patten and Walker............ Brackett and Haskins, report on nitrogen........ oh ste des Pnsler en Carbonates, in soils, report by Ames............. , pele re Berean 2 tts f Curate; ammonium, neutral; report by Walker. .3.-.....4--.+0c-¢-+2eree ence men UALS OS lyst yoouwe sede oD ecesHodaotns laceovenemEe cob Colors, recommendation by Mathewson................0.+-0-eeee ener tees Sape report by Mathewson. . Sort SONG ei A AO ta Pes in te Committees, appointment and = arsenal. Seog en ee es Definitions, food, report by codperative committee (Frear).................... Extract, ginger, paper by Harrison and Sullivan................ Extracts, flavoring, recommendations by Paul.............. Bei ig hear eens TEPOLe by; Ealls-eeeaeerceee SAE a Rene Pena Se TCS OE Hatsrancios, recommendations) by, Mermececa. 42-4) 06 fee mes le adeeenee acres POD OLG) YAEL Dregs ystsveree cree tar eins 8 thes ase Fialasaieeudtodenaters od nerely asm aievere Flavoring extracts, recommendations by Paul................-...+--+--+-0-ee: MEPONUMD VM AUVs peer tee ia acer Metay ulead ateh m asuck ovehe i ciowe mene eceveetheas Food, adulteration, report by Hortvet...... 2 BRON eR eH pu aan TAA Oe ee are te definitions, report by codéperative committee (Frear)................... standards, report: by commiuttee (Hrear)).........0.<.snec+seeclner seer ac Fraps, paper on interpretation of soil analyses......:...........-.---.0e-e+eees Frear, report by committee on food standards..................-.222.000000005 report by coéperative committee on food definitions.................... Fruit products, recommendations by Gore............... 000000 e cece eevee ees FEPOTUDYAG OLE sc chi Name ele isi echehine Ss Stason REE Ginger extract, paper by Harrison and Sullivan........................000-00- oodnow, TepOLbiONy VIDELAT a. ciao saci eee ea ee eRe ee AC ONE PeEDOLbVONTILULL PLOGUCEB Renu eta len minis tk ict ep ec Eee ae els oer eee 528 INDEX Halipaperonetrammonium citrate... aaeeeeer cere cece nee eee eee 375 Hare spaperon al kalisinosoils 2.).ssie.s am «/scis)--- esses eee 506 Hartman yreport OM: WIIG i/ssxc.).,s:c-Sos crs, care aerhero estate rd as Vee oe Seti 485 Haskinsjand Brackett, reportson mitrozenw-s.peere eer sernit ice eset 380 Ents report” On SPICES? ance ese isin cieisteloecun en ne Oe ee eee 510 Honey; report by Shannon's eerie He seve ele es lee eee cee ee eee 472 Hortvet. report on food adulterationse esse cnaetmondiseser erect eee nneere 465 Insecticides, recommendations’ by Roark. .-..5......2.<-- «0-6. -«+2 scien sss eneee 456 report: by: (Roarke sae nese ence tisveeorer- tise Ree eee ere 435 Jarrell; report on) determination of potashs...5.2e.. | 22.34 17.37 | 17.23 G. G. Parkin.......... 1n.o2 |{ 4-4 |} 15.49 | 15.34 a1.s2 |{ 32-37 Ua ees 1916] HORTVET: DAIRY PRODUCTS 543 It should be recalled in this connection that the results reported by a smaller number of collaborators in 1913, on sweet cream, homogenized cream and ice cream were in general quite favorable; and for purpose of comparison at the present time these results are herewith included. Many of the collaborators who have submitted reports during the pres- ent year have had experience with similar work in years past and some are included among the collaborators of 1913. As already stated, all or nearly all collaborators are experienced analysts and represent eight well- known laboratories. A critical examination of the tabulated results tends to the conclusion that there are peculiar difficulties attending the determination of fat in evaporated and condensed milk—difficulties which possibly have not heretofore been fully realized. The collaborators, doubtless, in some instances reported results which should have been omitted, that is, some results may have been reported which were recog- nized as faulty and represent first-trial attempts. These, however, are included in the tabulation, as no definite information has been obtained concerning the relative merits of the different results submitted. In the main the results obtained by the continuous extraction method, both on the samples of sweetened condensed and unsweetened evaporated milk, are decidedly lower by the continuous extraction method than by the Roese-Gottlieb method. An inspection of the results cannot fail to reveal the probable composi- tion of the samples prepared by the manufacturer. It is a very safe prediction that the samples of unsweetened evaporated milk were made to contain not less than 7.8 per cent of butter fat. Nevertheless, there are twelve results obtained by the Roese-Gottlieb method which are so seriously low as to attract attention. A similar state of affairs is revealed among the results obtained on the samples of sweetened condensed milk. There are evidences that each collaborator has, however, succeeded in obtaining fair check results in his own laboratory. The great discrepancies appear to occur among the results obtained by different individuals. Per- sonally, the associate referee has full confidence in the Roese-Gottlieb method as a reliable procedure for the determination of fat in evaporated milk, and the same opinion will doubtless be expressed by every analyst who has'had considerable experience in the analysis of products of this kind. It is unlikely that the various collaborators deviated materially from the procedure as outlined in the directions given, yet, just where the difficulty hes it is impossible to point out. The conclusion seems inevitable that there is here indicated an imperative demand for a special investigation into the details of the various methods for determining fat percentages in condensed and evaporated milk. It is quite likely that the difficulties are not serious, but it is at the same time plain that they exist, and that there is somewhere a lack of adequate understanding among 544 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 4 analysts respecting essential features which may be peculiar to such prod- ucts as have been considered during the past season. Two collaborators who reported that the samples of evaporated milk had changed somewhat in shipment, stated also that no great difficulty was experienced in preparing the samples in proper condition for analysis. It is too early to conclude that the continuous extraction method is in- capable of yielding at least fairly accurate results; on the other hand, it appears to be well worth while that a further attempt be made to cor- rect deficiencies and to modify the method with a view of clearing up a number of obscure points. Also, in view of the general application of the Roese-Gottlieb method, it appears to be seriously demanded that this association recommend that special attention be given to this method, with a view to correcting deficiencies if any such actually exist. While the results obtained in 1913 were far more favorable, although reported by a smaller number of collaborators, no recommendation can be made in favor of the final adoption of the continuous extraction method at this meeting. The situation which now seems most worthy of consideration is the status of the Roese-Gottlieb method applied to evaporated and condensed milk, and it is therefore urgently brought to the attention of this association that some action be taken with a view of the study of this or any one or two other methods that may be proposed. Whether it is possible to expect much more uniform results than are shown herewith among collaborators working in different parts of the country under differ- ent conditions and possibly on samples arriving over long distances, it is not easy to predict. There is also always the chance that samples are not strictly uniform in composition. Nevertheless to any one who has had daily experience in the analysis of these products during the past few years, taken in connection with the general showing presented in this report, it must seem imperative that this association give some serious attention during the coming year to a study of methods for the deter- mination of fat in sweetened condensed milk and unsweetened evaporated milk. C. W. Bradbury of the Chemical Laboratory of the Virginia Depart- ment of Agriculture at Richmond, read a paper on “The Alkali Method for the Determination of Fat in Ice Cream and Condensed Milk,” the essential facts of which had been published as Circular No. 42, of the Dairy and Food Division of Virginia. No report was made by the associate referee on cereal products. 1916] MAGRUDER: VEGETABLES 545 REPORT ON VEGETABLES. By E. W. Macruper (Department of Agriculture, Richmond, Va.), Associate Referee. CANNED FOODS. Last year the work on vegetables consisted in determining the per cent of easily-separable fluid found in canned tomatoes. In that report the results of the examination of 77 cans of tomatoes were given and the average amount of easily-separable fluid was found to be 52 per cent with an extreme variation of from 38 per cent to 64.4 per cent and with 60 per cent of the samples ranging between 47 per cent and 57 per cent of fluid. It was recommended that the work be continued and that other canned foods be included in the study. The work recorded in this report has been done by J. B. Robb and the associate referee and has consisted in the examination of canned tomatoes, peas, and lima beans. Not only has the amount of easily-separable fluid been determined, but quite a detailed study of the fluid has been made, consisting in the determination of the specific gravity, the solids in the unfiltered and filtered liquid, the immersion refraction index, and in some cases the polarization. An endeavor was made to determine whether there was any constant relation or factor existing between the solids in the filtered liquid and the refractometer reading. METHODS. Separable fluid.—The method used was practically the same as that employed last year: Determine the weight of the contents of the can and transfer the ma- terial to a regular fertilizer sieve with round holes 1 mm. in diameter and stir the material gently with a spatula to allow the liquid to drain out, allowing the material to stay on the sieve 5 minutes in the case of tomatoes, and 3 minutes in the case of lima beans and peas, stirring gently just before the expiration of the time limit. Some liquid still remains with the canned material at the end of the time specified, but the great bulk of the fluid had drained away and the time specified is considered about right to allow the easily-separable fluid to drain off. This fluid is then weighed and the per cent of easily-separable fluid calculated. Measuring the contents of the cans and the fluid was tried and the results were about the same as obtained by weighing, but on the whole weighing was found to be the easiest and quickest. Solids.—Dry 25 cc. of this fluid on a steam bath and then in an electric oven for 3 hours at a temperature of 105° C., and calculate the per cent of solids. Filter a portion of this fluid and dry 25 cc. of the filtrate exactly as in the case of the un- filtered fluid and calculate the per cent of solids. Refraction and polarization readings.—Use a portion of the filtered liquid at a temperature of 20° C. for the determination of both the refraction and polarization readings. The immersion refractometer and Schmidt and Hentch polariscope and a 100 min. tube were used. In many cases the liquid was so opaque that the polarization could not be read, so that few results are given. 546 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. J, No. 4 Factor.—It was hoped that a factor of value might be obtained and in order to _ obtain one the refractometer reading was divided by the per cent of the solids in the filtered liquid. The results are given in the tables. Five brands of canned tomatoes and two each of peas and lima beans were selected and a case of each was obtained. Three cans of each brand were examined in December, 1913; then three cans of each were packed in a case and shipped by express to W. F. Hand in Mississippi, who im- mediately shipped them back, thus giving th2m a journey of about 2,000 miles. After their travel they were examined in January, 1914. This was done in order to determine what effect a long journey on the train with the jolting incident thereto would have on the contents of the cans. Three cases of each brand were kept without being moved and examined in July, 1914, which was seven months after the first examination and about one year after they had been packed. This was done to determine what effect age had upon the contents of the cans. Some other cans of beans and tomatoes were examined in the fall of 1913, and the results are given in the tables. There were also examined in October, 1914, samples of tomatoes of the pack of 1914 of two of the brands which had been examined previously. RESULTS ON TOMATOES Easily-separable fiuid—62 cans of tomatoes were examined and the average amount of easily-separable fluid found to be 50.7 per cent with a minimum of 31 per cent and a maximum of 66 per cent, or a difference of 35 per cent, which is greater than the lowest per cent. A striking fact shown is the very wide variation between cans of the same brand examined at the same time. This is shown in the case of the Cherokee brand, which varied from 31 per cent to 53 per cent with a difference of 23 per cent, and again varied from 39 per cent to 60 per cent with a difference of 21 per cent. Only about 100 cans of this brand were put up by a farmer for his own use and not for sale. They were usually put up in batches of about a dozen at a time, which may account for the wide variation. No water was introduced into the cans. In comparing the first results obtained with those after shipping and after standing it will be seen that shipping had practically no effect on the amount of fluid, but that age seems to increase the amount of fluid slightly. Some of the cans from the same cases are still on hand and we expect to examine them during next winter to see what effect six months more of age will have. Specific gravity—The specific gravity does not seem to bear any re- lation at all to the per cent of fluid, as some samples with a high per cent of fluid have a high specific gravity, while others with just as high per cent of fluid have a low specific gravity and vice versa. Solids—The solids in the unfiltered and the filtered liquid bear a close relation to each other. As a rule the unfiltered fluid has the most solids, teh |") B98 | 69°F | 92°F | EZOT | 2°09 | “st1Nser [Te Jo osBIOAW 61°L Jk 1— | 888 | G89 | L8°F | ScOL | Soh |BA ‘AunoD pyeysoyseyO | «doy [ITH | Z1Z9 19°Z |S°I— | &°88 | 06°F | 86°F | F20T | 8'IS “aA ‘AqunoD pleyseysoyD |-dor tH | 19 gL |6°I— | $68 | FL°S | 20'S | G2OL | &' FF | BA ‘AJuNOD pregseyseyO | “doy TMH | ZIZ9 8g°2 |0 6'6f | $2°9 | Z's | 9ZOT | Z'69 | "VA ‘AuUNOD pyegieqsoyH | “eaHoIyD | ETZ9 LiL |€ £°0F | 02° | GTS | LzoT | $89 | BA ‘AgunoD pragreyseyH | “eexo10yD | €1z9 2G°L |Z 9°68 | 242°S | 20S | 9Z0T | O'2g | "BA ‘AQuNOD preyszeyseyO | “Payor | E1Z9 ve qu a9 id ‘ODBIOAT OY} UL POSN 4OU Sz[Nsei OS ‘OpLUI uOIZwUIWIOJap a10Jaq poyUoUIIE; pinbry a, Lae AI = FOL |" | FE | GFF | SOF | ECOL | SSF |°*°'89[Ns01 aquIo.0q []# Jo esvIOAY se| 35 sce =| a3 ae E é £18 | 20'% | S6°S | OZOT | a FS Be] ss = Fees || Hen || io) : i GHG | BA ‘AgunoD syouwoy |" WeIsZOUO0TY |0L EES BE ga: eee || teen || eee] FE 9 €°G8 | STF | STF | OZOT | 2°09 |'""-BA ‘Agun0D S1ouEON “cere BOUO aro er eS lee &.. | sn) 22 5 8'TE | TOF | 16°€ | OZOT | 9°8¢ |'-"-eA ‘AjzUNOD exouvOY |*"*UTeIZOUOT |0LT9 Se IDES |e soil) es = Namvao ay 2 8'SE | 6E°S | FE°o | OZOT | F'ES BA ‘AJUNOD ayxouTOYy wavasou0Wy |6919 aS7) 15 lege ¢ I INVaa g i : T aas| t° [68 g 3 5 1'@e | 96° | 46°E | GIOT | 6°99 |* "BA ‘AzuN0D exouvoY | **uIBIZOMOTT |6919 Sea Es (eo 4| 5 3, 5 8°98 | 26°F | 86°F | SOI | FOF |°-""GA ‘AJuN0D oyouvoy |" "weISOUOTY |69T9 SAe| eB |pQg) = = 3 £08.) 268 | Zhe | SIOL | TLE |°BA ‘AgunoD pyeysoysoyO |" "dor [tH [2129 woo] Flome| 8 5 x O'TE | G's | GP'S | STOT | Leh [BA ‘Aqun0D praysoysoyO |°* "doy, [TH 2129 “s| a |SrF & & 2° T'Ge | #L°E | 9L°E | BIOL | 6'6F |"BA ‘AZuNOD pyegseyseyO “dol TH |2129 9°E | 06'E | LIF | TOE | Teh | BA ‘puoumyory |***"uMoysourEl |FOT9 “HT6T ‘4990790 UA paununrgy “FIT fo Yq PRE OO LEP LEO ee rca EA ave |e eee 0O'2 |"""""|e°Be | €8'F | 98°F | Ezor | sec | TA'2 | Z'98 | 12°F | 28°F | ecor | 9709 | 64 4 GLE | 08°F | 06°F | FOOT | £6) |" oduIOAW 90°2 Jo'c— |979e | 2T°S | co's | ezor Jog'se | es°2 |4T— | Ese | 1o°F | SovH | Teor | FTE | SE 4 £°8E | 12'S | 02'S | OZOT | F'6E |'BA ‘AQuNOD pjoysezsoyO |" *** “-eexoISYO |£1C9 De Tee eae fa 08 BF 60°8 |6'1— | 6&6 | OIF | 08°F | OcOr 1°88 Om 9'0F | 69° | €9'S | 8ZOT | €°6F |-BA ‘AqunoD pregreysoyO “1s gafou0t|3 £129 OS AUB | Bats sey 29°L \T'G— | 8°86 | 90'S | LE°¢ | OcOr | B’es | Fe Z'0F | FES | ES°S | BLOT \00°09 BA ‘AquNOD plegaoysayD |**-"-eoyorayD [E129 AIDE Bea Pa Gray aLEDG|Y "99 | 22° L if gc | 2o°F | 90°S | Z2OI | 8°6¢ re T'€ | 60°F | €2°F | 2201 | O'LF “PIN e1ougTeg |jOryO purpAreyy 9619 =H MI Dareea bear eal eCUay | CeO vente 28 2 | T— | O'€e | OL | Te" | Teor | 2°99 | OF 8 FE | Gh'F | OOF | EzOr | z ies “PIN (exourry eg Fee) purpsreyy}9619 Se APS | | SRE OER ED |e a] Oe | Me es Beg oo ea Sp |B | Her | bap | oo oaa snout (ea a 9 ‘lr6er | 68°F | seh | Ozor |e8°9¢ | 99°2 |o°z— | ovce | 2o°% | 29°F | coor jos'es | S€°8 6'E& | 90°F | TGF | GaOT | S°aF | “soem sSuOBIOMLEAY | 3019 Ger | €8°h | FEF | OZOT |2L°2o | 18°8 |¢'s— | GFE | OZ F | Zot | GZOr |O6'Tg | 09'8 Z'86 | 98'€ | £0'F | TZOT | OOF | ‘STOSION IM S89T9 O'TEr | €8°F | 82°F | GIOT [FF GS | 2O'L | BOF | Of | EFS | LEOT | S'oF | BEL 6'6E | ¥F'S | GE'S | 9ZOT | L'OF |---77 7 7 8A ‘ouRO, L919 O'ger | 29°S | 29'S | Gzor |go°zF | O9'2 |e's— | GRE | 219 | OTs | Le0r | T's | 994 £'LE | 98°F | 68°F | EZOT /00'8F |” “BA ‘ouBoy, L919 “=-lo-Ter | @t'F | G&S | FOT eZ eh | €9°4 |o-— | 9'6e | GIG | Ges | Leor | STH | 982 “+ | 6 68 | 82°¢ | GES | 920I j00 se Nes B A OLY OT: 2919 "*-I¢-er | 00'S | STS | FzOr |8e'09 | SIL |F'Z— | 8°ZE | 92's | EFS | EZOr | O's | S82 ¥'8E | 06°F | O'S | 9Z0T | Zag ‘BA ‘BIngsyowepery 9919 ‘|oger | 12°¢ | 82'S | EOE [FE so | FOL |I's— | e'SE | cop | 8h | Geor | e789 | O82 | 6 | c0°¢ | 6E:9 | 9zOr joo 0g | ° BA ‘Banqsyorsoposyy 9919 acer | 08°F | 06° | ezor loF'z9 | #9°2 le s— | F2e | 96°F | SIS | gor | O'ZG | TL'2 |'-*°"*| T'6E | Loe | E'S | LZOE Joo'os °° “VA ‘BInqexorrepory 9919 quaa | 2uao 2ua92 qua | 7Uaa quad quas | qWa9 7uao sad dad. Jad wad dad Jad dad wad wad lanl na mM foal U nm id | yh nee ag sce Be mer = E nee ay eae =, & oo E nee may +-8 eee g 2 Ss 3S a 5 Be] BE p32] 2 A= ttlac| fe teu tse = ep lees! se | & | 2 ; ees| Se lasel2e|26|ee| & |ges| £2 |ess|25| & | 25) & \g63) Se aae(G2| & 28] & 9 QR) oe |eeg| & [ae | ee] & as ee Rogie |e fae] Fgh ee Pega |e pee) a = ae a ia F=c=itsy| een ele Feuehr|| Test es 5 gfelBy ee| ©] 2| 3| [eGelbs ieee] #| =| 9| |efele ieee] | F | § g SS.0) + aM 5 =5.0 E.% 2 2 ishesller E. 2 = aS gn |B =F < Beem) ety 1S =] ee et A €. £ £ age ce Beg, = ee & ree ce 228 = 2 E eee ee 2o3 2 es e NOILYSO'L aNvua 2) os —. Noe 5 a =e) e a ot eet 0 2 g Z3| &\&FF] 2] 2] 9 S3| & SPF] 2B 2 Z3| ge lSPe| & 2 FIGI ‘A7TO8Lr ‘XUVONVE GANINVXG ‘ GUNIWVXG ONY aoVv1d GNO NI caaH eRe 000'2 SS4uaXaG Xa agaarHs EUG ete CHIC D 1 Acs *S20]DUL0? PIUUD) CIR A 548 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 4 but in quite a number of cases the reverse is true. This may be caused in part by the evaporation of some of the water during filtration, as the liquid filtered very slowly, but to reduce evaporation to a minimum the filtrate was caught in narrow-neck bottles. It is rather surprising to find that the filtered liquid contains about the same amount of solids as the unfiltered, the average amount of solids for all’of the tomato samples being 4.76 per cent for the unfiltered fluid and 4.69 per cent for the filtered liquid. Refractometer reading.—The results of the refractometer reading are quite interesting. They vary on the whole about as the percentage of solids, but here as in all the other determinations we have erratic results. The readings marked with a superior ‘‘!’’ were made on liquid which had fermented to some extent before they finished filtering, hence are lower than they should be, and are left out of the general average. On the whole the refractometer reading is more characteristic than any of the deter- minations, and as it is one of the easiest made it is probable that it furnished a better indication of the character of the liquid than any of the determinations. Factor —The factor obtained by dividing the refraction reading by the solids from the filtered liquid is not at all a constant one, although the variations are not as great as most of the other results obtained. It is doubtful if it furnishes any information of value. Shipping does not seem to have any appreciable effect on tomatoes; age has a slight effect on the percentage of fluid, but apparently none on the composition of the liquid. RESULTS ON CANNED GREEN PEAS. Examination of fifteen cans of green peas was made and the results are given in Table 2. In the case of peas neither shipment nor age seems to have had any appreciable effect, as the results are about the same throughout. There is very much less variation in all the determinations than in the de- terminations on the tomatoes. The average amount of fluid was 34.9 per cent, with a variation between 28.7 per cent and 38.6 per cent. The solids in the unfiltered fluid are rather erratic, while the solids in the filtered liquid are quite uniform. The refractometer reading is very uniform, varying between the narrow limits of 38.6 per cent and 42.8 per cent. The factor is also quite uniform. RESULTS ON CANNED LIMA BEANS. Examination of fifteen cans of lima beans was made corresponding to the fifteen cans of peas. In addition six cans of a rather miscellaneous lot of beans were examined, some having been soaked; the results ob- ve[ecreee|eceee|ecteelisesee[eceeeeleetee[ersers|eceeeeleeeeee[ereeceleeeeee[eeseeeleessest ge-g | g-el gp logcglegz |--:| g-te |occ777 7: eseroay “eo €e°L | 6°€| 3°98 “PW ‘odourTz[eg |" **" “"BUITT PexBOS |S619 aS OIL | fF OLE “PW ‘sourry[eg jo" "*"*"***" "BUILT pexBos |S619 os 26'¢ | 2'&| 18h “PIN ‘orowra[eq |"*'*'(payBog) BUILT woysog |F619 <6 06°S | GF) 2 8F “PIN ‘e10urry [eq ‘(pexBos) BUTT uosog |F619 2 0L°9 |'"*"| 9°68 “BA ‘puourpony | "“suBeg 1093nNgq UlBIZOUOPW | E819 3° OF'9 ¥ S| Vit pice A ‘puoulyory ““suveg J033ng wWBIZ0UO]Y esto S0n.00| [opes05|jos0000 lbonon|joq5000 veseesdoveeee] porg Aonncal sanstoooaon|leboo) lace veoloesee) ogpxoan parane5 T6308 |e PL SBn eS LiGa Gedy seen ONOe a nObEO, 09°9 |°***| B&F j¢2°9/8°8 j98OT] gee |" esBIEAW 20°9 90°9 | 9°2| 9°oh |Fe°L/99°OT |OFOT) se | “LN ‘A3tO Aessor “Buos |STZ9 98°¢ 19°S | Lb) #99 [LZ G/EF TT |TSOT] 9°FE | “CN ‘A910 Aosser j “euog |STZ9 6F 9 92°9 | FS) 2 °Sh |ZZ°2/2% OT \680T|00'SE | “CN ‘A310 Aossor . “Buog |S1Z9 ¥6°9 68°9 |""*'| 0°88 |TS’S|S6"¢ |TEOT|00'OE | "BA ‘puoUrToRy - dB9 OTM |ZLT9 &o°L &Z°L |°°""| 8°98 |60°S/GE°S |4Z0T|00'SE |°*" "BA ‘ puoUTTor Tree OMG |ZLTO 80°2 16°9 GSE |29°9/96°S jOBOT\O8 TE | “BA “puouryory seeds OITA [2279 | “suDaq DULY) paUuUuny)y ‘¢ ATAVL IEEQY | os 8°0F | 90°9 | 0&2 | O£OT | 6B FE | corees}eresdesss**/Q9BI8AB [BIOUOD) CSA ihe GIF | 88°9 | 89°8 | SOT | 8 FE | OS 9 €1°L |O80T| 9°#E | OdBIOAY 06°2 "| #68 | FPS /68°OT | SeOT | 2°8% | L¥'9 28°6 |Se0r! €°08 |°" BA ‘puouryory “UBpByMod |FIZI 89°9 "| €°GF | €8°9 |80°OT | GEOT | F SE | 99°9 06°2 |6Z0T} 9°FE |" "BA | puouAyory UBBYMOT |F1Z9 98°9 €'1F | 9T'9 |249°8 | SeOl | 9 FE | SE9 G98 |Z0T) 9°08 | BA ‘puourqory UByuyMod |F1Zo Té'9 |" €°@h | 18°9 |#o'L | TOT | Tse | e9°9 88°9 |2201}00'8E |*""A “N ‘10989q00R exnd Avy |TLT9 90°9 8°GF | 90°2 |Z9°L | Of0T | 9:98 | 69°9 29°9 |8ZOT|T8°ZE |" "AN (1e3seqooy eyng Avy |IZT9 FE'9 OIF | 9F'9 |0@'L | TEOT | 9°88 | 6F'9 89°9 |620T/€9°SE | "A "N “qoqseTOOU, “eynd Avy |TLT9 quad | quar quao 299 | Quad 7499 739 399 dad | sad dad wad | sad dad wad sad re) wml m pal otlecbl et] 2 let] EF leek] oF laebl oF) Flas] EF le kloclen Bla! 2 [8] & 3 SS8l se (Seb) oe] & (a8!) € |Sb3] SP ISSB/ oe) & 1238] € |SBSISEISSBIOS| & [S8! € : S25| Se \gcslee] s |e) & legs) Se \ees\ee| & | ee] & lses/Seigeciet) = jee E g BRR gg leegl Ee | ep [ae] © Re ee gla | Eg | Be RP lagleegla | 5 [ae 4 A Salles a Ss e 2. EB 5 - BI ao, S|FES Hl = > see tlee| (| 2 | 3 gepjes ieee) Fle] 8 gegeelea’| *| & | 3 g aed) og |o th . L Ram oa |e Feu) iG: < rams alo 5 JOLLY 7 ‘al £28 a 288 = z & 5e8 cB B28 = Re Es Re8 ehle33 | a ES NOILYDO1 aNnvuad 5 B) ee |p 5 S) = load 5 g..— =] 4 f3| a iSFe} & 2 3| a |SFo| & 2 | g/S72| 8 g FIGI ‘ATOL FIGI ‘AUVONVE GANINVXE 161 ‘WHaNGOGG GANINVXE | aaUNINVXG GNY GOV1d GNO NI dday ANY SH1IN 000'Z S8AUdxXa AM Add HS | ‘spad pauun,) % WIAvL 549 550 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 4 tained on these beans are given in Table 3 but not included in the general average. The average amount of fluid was 35.8 per cent, with a variation between 30 per cent and 46.1 per cent. Shipment and age both seem to have had a slight effect in increasing the amount of fluid, but the increase is of little consequence. All of the results are quite wide apart, except the factor which is nearer a constant than in the case of the tomatoes and the peas. There were not enough brands of peas or beans examined to find out just how variable the results would be. The contents of all of the cans ex- amined were in good condition, and we believe honestly packed. GENERAL CONCLUSIONS. (1) Shipment does not seem to have any appreciable effect on the amount of fluid of canned tomatoes and green peas and a very slight effect on the amount of fluid of lima beans. (2) Age seems to increase the amount of fluid in canned tomatoes and lima beans very slightly, but hardly at all in canned green peas. (3) The per cent of easily-separable fluid is a very poor index of the character of canned tomatoes, peas, or lima beans. (4) The refraction index is probably the best index of the character of the fluid in all canned goods. A paper on the ‘Characteristics of Common and Lima Beans” was read by Arno Viehoever of the Bureau of Chemistry. REPORT ON COCOA AND COCOA PRODUCTS. By H. C. Lytueor (State Department of Health, Boston, Mass.), Associate Referee. At the meeting two years ago the Baier and Neumann method for the determination of case was recommended for provisional adoption by Mr. Dubois, associate referee. At-the meeting last year the present associate referee recommended the method for final adoption as pro- visional. Objection was made to this method by Mr. Dubois because he had found a sample alleged to be milk chocolate which gave no reaction for casein by the above method, but which contained the proper proportion of butter fat and milk sugar. According to information received from the manufacturer, this chocolate was made by a method somewhat different from that usually employed, and the committee on recommendations recommended further study upon the method before adoption as pro- visional. In conversation with Mr. Dubois he stated, in answer to a direct question, that the department had no knowledge whether or not the manufacturer used butter and glucose to simulate the chemical com- 1916] LYTHGOE: COCOA AND COCOA PRODUCTS 551 position of milk chocolate, but a very careful survey of the factory made after the manufacturer was given a hearing gave no indication that such was the practice. The Bureau of Chemistry intended to make a study of the condition under which this particular milk chocolate was manufactured and samples of the product were to be forwarded to the associate referee. No samples have been received to date and on October 28 I learned from Mr. Dubois that the investigation had been dropped by the Bureau. The associate referee has used this method for several years, not only for the determination of casein in milk chocolate, but also in other food products in which milk or casein is used as an ingredient, and has never experienced any trouble with it, but considering the objections given by Mr. Dubois, does not consider it advisable that the method be adopted at present as provisional. For the past three years the method of Ulrich! for the determination of cocoa red has been used in the Laboratory of Food and Drug In- spection of the Massachusetts State Board of Health for the detection of added shells. This method possesses no material advantages over the determination of pentosans or of fiber for this purpose, but as the product determined is a constituent of the cocoa nibs and is absent in the shell, while the pentosans and fiber are constituents of the shells and not of the nibs, the determination may be of interest to the association. The method is as follows: To 1 gram of fat-free dry substance, which should be finely powdered, in a 300 ce. Drlenmeyer flask, add 120 ce. of pure acetic acid (50 to 51 per cent); connect with a return-flow condenser and boil for 3 hours; cool and bring the contents of the flask to a volume of 150 cc. with cold water; shake well and allow to stand at least 12 hours; filter through a dry filter and treat 135 cc. of the filtrate (corresponding to 0.9 gram of the original substance) with 5 ec. of concentrated hydrochloric acid and 20 ce. of a 20 per cent ferric chlorid solution. Connect with a return-flow condenser, heat to the boiling point and boil 10 minutes; cool quickly and transfer to a beaker; after letting stand at least 6 hours, filter upon a weighed filter, wash- ing the precipitate with hot water until free from iron; dry for 6 hours at 105° C. and weigh. Results of analyses of 18 samples of commercial cocoa. (Per cent.) FAT-FREE SUBSTANCE MOISTURE FAT eee Ash Fiber Pentosans Highest 4.80 25.07 7.78 8.32 4.55 16.22 MOWER USs os cone eee 2.60 15.07 5.80 5.33 3.97 isl sey PACT ARE Mr cretien cee Uae 3.87 19.95 6.58 6.35 4.44 14.06 Dutch process, average. . 4.76 24.37 9 87 6.24 4.47 13.47 1 Der Nachweis von Schalen im Kakao und in seinen Preparaten von Diplom-Ingeneur Chrostoph Ulrich. Dissertation, Detmold 1911. 552 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, Nu. 4 The examination of 41 samples of commercial cocoa in 1914 gave from 11.22 to 17 per cent, average 15.17 per cent of cocoa red. Ulrich reports for different varieties of cocoa nibs averages from 11.12 per cent to 16.54 per cent. RECOMMENDATION. It is recommended— That the associate referee on cocoa products for the year 1915 make a study of the manufacture of milk chocolate with the view of finding out whether or not the casein is rendered insoluble in the reagents by differ- ent methods of manufacture. REPORT ON TEA AND COFFEE. By J. M. Barter (Agricultural Experiment Station, Orono, Me.), Associate Referee. According to the recommendations adopted by the association at its last meeting the work on tea and coffee has been confined to methods for the determination of caffein. One sample each of tea and coffee was sent out to all chemists signify- ing a desire to coéperate in the work, together with the following in- structions: INSTRUCTIONS TO COLLABORATORS. Determine caffein or thein in each of the samples by the following methods: Method 1.—Carefully weigh 10 grams in No. 60 powder into a 500 ec. Erlenmeyer, add 100 ee. of water and 10 cc. of 10 per cent hydrochloric acid and heat to boiling with reflex condenser for 2 hours. Cool, decant liquid through a filter, treat solid material with 3 portions of 50 cc. each of boiling water, filtering through same paper as above and then wash material on filter with 50 ce. of boiling water. Concentrate to 150 ee. by evaporating over steam or water bath. Transfer filtrate to a 500 cc. sepa- rator, Squibb type, add 5 ee. of stronger ammonium hydroxid, shake out with 50 ce. portions of chloroform five times. After the first shaking out let the separator rest until separation is as complete as it will be; then run chloroform into another 500 ¢c. separator; add the second portion of chloroform and shake again, after standing until no further separation occurs, run the solvent and the adhering emulsion, if any, into the second separator but do not run in any of the non-emulsified liquid. Re- peat three times, running chloroform and any emulsion into the second separator. Then discard the liquid in the first separator and give a second separator contain- ing the chloroform and emulsion a violent shaking; let stand and then run chloroform into a 250 cc. Erlenmeyer flask. If there is an emulsion remaining in the separator, add 1 to 5 ce. of 94 per cent alcohol and shake. When the chloroform has separated add it to that in the 250 ce. flask. Add about 25 cc. of chloroform to the aqueous alcoholic layer in the separator and agitate; after separation run the chloroform into the 250 ec. flask and then evaporate off the chloroform on the steam bath using a moderate blast of air and removing from bath as last portions evaporate to avoid ' Read by H. H. Hanson. 1916] BARTLETT: TEA AND COFFEE 553 spattering. When the residue of crude caffein is dry, add 10 cc. of dilute 10 per cent hydrochloric acid and 50 cc. of water and warm until caffein is dissolved. Cool and precipitate with 50 ce. of iodin solution (10 grams of iodin, 20 grams of po- tassium iodid, 100 ec. of water), stopper flask with cork and let stand overnight. Filter through 9 em. filter and refilter filtrate, if necessary, washing flask and pre- cipitate twice with iodin solution, but not attempting to remove all of the precipi- tate from the flask. Transfer filters to flask in which precipitation was made, add 0.5 gram of sodium acid sulphite or sodium sulphite, 3 ce. of 10 per cent sulphuric acid and 15 ec. of water and warm until iodid is decomposed, more salt being added if the amount is insufficient to decolorize. Filter into a separator (small 100 ce.), add excess ammonium hydroxid stronger and shake out five times with 15 ce. portions of chlorform. Wash the combined chloroform extracts with water which is discarded and then concentrated to 10 to 15 ce. add dry animal charcoal, shake and allow to stand 1 hour with occasional shaking, filter through a small filter into a tared dish, washing flask and filter three or four times with 5 ce. portions of chloroform. Evaporate chloroform and dry residue in a desiccator and weigh. Method 2 (Gorter method for coffee).—Moisten 11 grams of finely-powdered coffee with 3 ce. of water, allow to stand for half an hour, and extract for 3 hours in a Soxh- let extractor with chloroform. Evaporate the extract, treat residue of fat and caf- fein with hot water, filter through a cotton plug and moisten filter paper, and wash with hot water. Make up the filtrate and washings to 55 ce., pipette off 50 ce. and extract four times with chloroform. Evaporate this chloroform extract in a tared flask and dry the caffein at 100°C. and weigh. Transfer residue to Kjeldahl flask with a small amount of hot water and determine nitrogen by Kjeldahl or Gun- ning method. Nitrogen multiplied by 3.464 equals caffein. It is suggested that after shaking out the aqueous solution with chloroform, run the chloroform into a second separator and shake with a strong solution of sodium earbonate. The sodium carbonate solution will remove most of the coloring matter. Then pass the chloroform to a third separator and wash with water. Treat the remaining chloroform shakeouts from the aqueous solution in the same manner, passing them successively through the sodium carbonate and wash water. The washed chloroform extracts are then united, evaporated, dried and weighed. Method 8 (Modification of Stahlschmidt’s method for tea).—Boil 6 grams of finely- powdered tea in a flask with several successive portions of water for 10 minutes each, and make up the combined aqueous extracts thus obtained to about 550 ec. with water. Add 4 grams of powdered lead acetate to the decoction, then boil for 10 minutes, using a reflux condenser; add water so that the solution will finally be exact- ly 600 ce., and cool to room temperature. Then pour the solution upon a dry filter and evaporate 500 cc. of the filtrate, corresponding to 5 grams of the tea, to about 50 cc., and add enough sodium phosphate to precipitate the remaining lead. Filter the solution, and thoroughly wash the precipitate, the filtrate and washings being evaporated to about 40 ce. Finally extract the solution thus concentrated with chloroform in a separatory funnel at least four times and evaporate the chloroform extract to dryness, leaving the caffein, which is dried to constant weight at 75° C. and weighed. RESULTS REPORTED BY COLLABORATORS. Reports were received from only two collaborators; therefore, only a few results are given in the following table: 554. ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 4 Results obtained on tea and coffee. CAFFEIN SAMPLE AND ANALYST METHOD Weighed | ¢,om ‘nitrogen per cent per cent TEA: . “ H.H. Hanson, Orono, Me.| No.1. Fuller.............. 240° 3) eee No. 3. Stahlschmidt....... 3.09 2.74 INewamethod!ues-. sms cnc 3.38 2.81 Hi. @: Puller; Institute of || Nosi! Puller-s7...---....- 2:47 | 3 eee Industrial Research, | No.3. Stahlschmidt....... 2,61.) || ose Washington, D.C. COFFEE: Hs Hansonte-ese seers Norslombiullerseens te see eee 1.28: +)4. eee HNO s2s ae GROLGEL s4s.-8. See ae oe 1.36 1.04 NewnmethodSa-.--cetjocne 1 24 12h) He Crulleny ae. te eee NO silen BU Lettres rare sieteretas 1.12 |. eee Nts 2. COIR peoécoouedar 1:06 ||.) ccermeeree COMMENTS OF COLLABORATORS. H. H. Hanson: In preparing the tea or coffee for work we have found that the recommendation to grind the materials so as to pass a sixty-mesh sieve not neces- sary, and, in the case of coffee, rather impractical. Upon samples ground to pass a forty-mesh sieve exactly as good results were obtained, and, in the case of coffee, better results. With the Fuller method good results were apparently obtained with both tea and coffee, but it is a long and involved one, and, while apparently accurate, is time- consuming and involves very careful manipulation. The Stahlschmidt method for tea is apparently quite correct if the result obtained by weighing is not taken as a final result, but a nitrogen determination made upon this result and the caffein calculated from the nitrogen content. The Gorter method for coffee is apparently only approximately correct. In reporting the results obtained on tea and coffee I wish to call attentionto a new or combined method which has in our laboratory seemed to work out very well. This new method, besides the three given by the associate referee, has been carefully studied. The results reported are only those obtained after the methods had been tried once, so that the analyst was familiar with the method. The new method gave good results with both tea and coffee almost without exception. Some difficulty was experienced in first trying out the method with tea. In the mind of the analyst the new method is as accurate and considerably more rapid than the Fuller method for coffee, and the Stahlschmidt method with the modification mentioned is the quickest and most accurate method for tea, but only slightly more rapid than the new method. The new method, which is given below in detail, is taken in part from the other three methods. Place 5 grams of material in an extraction thimble and moisten with 5 to 7 ce. of water, and allow to stand for about 30 minutes. Place a loose cotton plug in the top of the thimble and extract with chloroform ina Soxhlet extractor for at least 3 hours. Evaporate off most of the chloroform and add 10 ce. of 10 per cent hydrochloric acid and 50 cc. of water, and heat to boiling to dissolve caffein. Cool, precipitate caf- fein with iodin solution as in the Fuller method, the solution being made with 10 1916] BARTLETT: TEA AND COFFEE 555 grams of iodin, 20 grams of potassium iodid and 100 ee. of water. Use 25 cc. of this solution for the precipitation, cork the flask, and allow to stand overnight. Filter this on a small filter paper (9 ce. is about right) and wash with some of the solution. Transfer the filter paper and contents back to the flask in which the precipitation was made, add 0.5 gram of sodium acid sulphite or sodium sulphite with 3 cc. of 10 per cent sulphuric acid and about 15 ce. of water. Warm this until the iodid is de- composed, cool, dilute, make ammoniacal and then make up to the mark. Filter through a dry filter and take a 50 cc. portion for the extraction, which is made with chloroform, using about 5 portions of 15 to 25 ce. each. Evaporate off the chloro- ‘form and determine the nitrogen. The nitrogen content multiplied by the factor 3.464 gives the caffein. Results obtained on tea and coffee. (Per cent.) RESULTS SAMPLE AND METHOD FINENESS OF SAMPLE Weiahed | Caleuated TEA: LA RIUTO. Ceres ent ec Ieee 20-mesh PAS (I ah all re aS Fd o:6 LAU is cay ae eio EE ee enter 60-mesh 2683 Fw cee ee Siablschmidti-. .... aig enh tei eae Apple juice’. . SR AG CERI Siteaete ar eae 0.006 N. Y. 42677 | Pineapple juice, = (Curis ee eee ee eae ae Sete 0.007 AZ099N | Strawberry liqueurs 5.0.2.4. 50 ene esac se ene: 0.003 43100 | Cherry sirup..... En 8 Re se Rh ee 0.006 4499 Wa Cranberry jar scies....ce0e «sero io swiaare-ts eoie eens = 0.018 44702 | Orange pulps wee be a Pore Pan Ao ree 0.002 45595 | Lime juice. . nts oe SIEM Lye eee oe 0.001 47072 | Plum jam.. ee ae CO Ceres Draft 0.005 45583 | Raspberry sirup . eden -Sorecp de mcce ono caee 0.005 45623 | Blackberry Cordial eee Tieeteey ACT OS eee 0.003 45859 | Red currant jelly.. Re singh. ye pete 0.007 45860 | Raspberry and currant jam. . iihcvarstele Une a doen 0.004 45861 | Raspberry jam. ly See SEO arin pete sees 0.011 AGHEO GI CId erm a ieedile aos chats Po iey eke seit : 0.002 46134 | Wild bramble and apple ey ores sath BAe 0.004 461330 |sOrangermarmaladels.cn)..... 0.2.2... +0. eeeale ae 0.003 46826 | Raspberry sirup . Ee ee debe: cinerea eae & 0.006 46827 | Strawberry sirup. aa Je SORT e Res Manic ere 0.008 JOSS UG Dimer Ur Ce tanya sae wa ac cet) telson atte oe 0.005 47009 | Lime juice cordial (contained sé vlieylic Cid) eas ee 0.006 AGU es i Cuxramigellya fn. tesadeey a es vas elas sya 0.013 47527 | Strawberry juice................. Betas aun aoe : 0.004 47652 | Lemon juice..... Rae rss 2 tht AAA eR et ae 0.003 47918 | Apricot pulp..... SN coe Pte ita: siete ceases ER ee 0.004 A807 6m | Chermevacordiall scmcee tee one. hens ote te ae ts 0.008 4821S Pin CApUlersinUpepe eens a clan c sacl eco daltons 0.005 48331 | Raspberry sirup....... = ARSE BA Ao ecore babe 0.006 A Hell VAS DOCLEVESILUD ebier iis somteice eiisis cnenistne eeion haan 0.005 48720 | Pineapple pulp (Straits eeu ements) ede eed saci 0.005 HOO0IIS | Apricotgpulp 2. retains. = arian solani omen erect 0.006 AOL 2mr RAS PDECEIVAT AIM —cse Antes deat eee nee nene 0.007 1 Pressed by associate reteree. 560 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 4 TABLE 2.—Continued. Analysis of fruit juices and commercial products. MERCURIC CHLORID SERIAL NO. SUBSTANCE ee FORMIC ACID per cent 49013 || Straw berrysjameeye ace acer ac ace ise 0.008 49014™)|/"Black#currantiy amin een tener ee see eee 0.008 49105: Cherry, jam pee® soos ocrcagaccerrsarst eke ee 0.004 49106)"|Strawbernyajamaceeeas ee eee oer coe one nore 0.009 49107 "||| ‘Raspbernyajambess see. onc euiten ae eee eee 0.007 49147) «| Strawhbernysjamierneent ccm orrcs 2s ate cies 0.010 49195, || -Apricotipulpaccmemesecs someone ne beeen ener ees 0.006 49198; || Apricotipulpate aqme cn oetree eck ali coher ects 0.004 49221 | Pineapple pulp (Straits Settlements)............. 0.007 49298 | Pineapple pulp (Straits Settlements)............. 0.004 495215 (Strawberry; pulpaeseeneaaereen eee eee eee 0.003 Various | Highteen samples of honey from Cuba, Santo Domingo and Mexico varied between........... 0.002and0.016 Among the miscellaneous samples examined were a Russian caviar (0.011 per cent formic acid), two samples of English table sauces (0.021 and 0,011 per cent), and a Japanese sake (0.001 per cent). The referee wishes to express his acknowledgments to M. G. Wolf of the New York Food and Drug Inspection Laboratory to whom credit is due for the examination of a large part of the commercial samples here given. The fact that honey naturally contains little or no formic acid is fur- ther established by the recently-published work! of several authors who have examined authentic samples and found between 0.001 and 0.028 per cent. During the past year five fruit products of foreign origin have been found to contain added formic acid. The amounts found ranged between 0.03 and 0.25 per cent. The products consisted of two raspberry sirups, a black currant sirup, an elderberry sirup, and a cranberry juice. In one case the manufacturer furnished the information that 0.05 per cent of formic acid had been added to his product. The amount found was 0.041 per cent which accords well with his statement when it is considered that the acid used by the manufacturer was probably not absolute but of a concentration ranging between 60 and 80 per cent. The results of experimental work upon a practical qualitative test for formic acid have proved somewhat unsatisfactory. A practical qualita- tive test should give an absolutely negative indication in the absence of added formic acid. The reduction of the formic acid in the steam dis- tillate of food products by means of magnesium ribbon according to the 1 Ber. d. Nahr. Unters., Bromberg, 1912, v. 20: Jahresber. d. Nahr.-Unters- Landwirtschaftskammer fiir Brandenburg, Berlin und Frankfort a. O., 1912, v. 34: Jahresber. d. Nahr. Unters., Kiel, 1912, v. 16: Jahresber. d. Chem. Unter. Hannover, 1912, v. 18. 1916] SEEKER: PRESERVATIVES 561 method proposed by Fincke and given in detail in the associate referee’s previous report was found in this year’s work to give a slight positive test in a number of samples of natural products in which added formic acid was known to be absent. Ié is true that when added formic acid amount- ing to about 0.1 per cent or more was present the test became decidedly stronger and to an experienced worker the results probably would not prove misleading. The associate referee does not feel that it ought to be recommended as a general test. The tests proposed by Shannon! are excellent and reliable. They are based upon four characteristics of formie acid: Its volatility with steam, the crystal form of its lead salt, the formation of carbon monoxid when heated with concentrated sul- phurie acid, and its reducing power. Two of these characteristics are employed in the Fincke quantitative method with less consumption of time and with the added information of quantity. The crystallographic recognition requires special knowledge, but the formation of carbon monoxid taken together with the results of the Fincke determination is sufficient and convincing proof. While Shannon’s method is excellent for obtaining this test the Wegner procedure? is perhaps to be preferred because with about the same amount of manipulation and time quan- titative results are obtained. Working with the apparatus recommended by Rohrig? the associate referee obtained the following results upon two pure fruit juices containing known amounts of formic acid: Formic acid in pure juices. Strawberry juice.......... 0.0992 0.0941 0.0982 (LOWE, ona ketene Ss eet 0.0496 0.0465 0.0500 RECOMMENDATIONS. In view of the excellent results obtained by the collaborators with the Fincke method last year and considering the results of the examination of the varied commercial and natural products during the present year, it is recommended that the Fincke method be adopted as a provisional method by the association, with all the details as given in the report of last year. It is also recommended that the Wegner method be submitted to trial as a confirmatory test, and that steps be taken to secure a reliable quanti- tative method for the determination of saccharin in foods. No report was made by the referee on water in foods and feeding stuffs. 1 J. Ind. Eng. Chem., 1912, 4: 526. 2 Zts. anal. Chem., 1903, 42: 427. *Zts. Nahr. Genussm., 1910, 19: 4. 562 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 4 REPORT ON INORGANIC PHOSPHORUS IN ANIMAL AND VEGETABLE SUBSTANCES. By E. B. Fores, Referee, and F. M. Brneie (Agricultural Experiment Station, Wooster, Ohio). INORGANIC PHOSPHORUS ESTIMATION IN ANIMAL TISSUES. In previous referee work on inorganic phosphorus estimation in ani- mal tissues three methods have been compared, namely, the neutral molybdate method of Emmett and Grindley, the modified barium chlorid method of Siegfried and Singewald, and the magnesia mixture method of Forbes and associates. Satisfactory comparisons of these methods have been made on muscle, the results being practically identical; and certain important limitations to the applicability of the two methods first mentioned, to tissues other than muscle, have been established. It was now desired to test, by the method of recovery of added phosphates, the accuracy of the magnesia mixture method, in its latest form, with animal tissues of diverse character, and also to study individually, a num- ber of details of this method, as: (1) The influence of heat, as specified; (2) the method of filtration; (3) the completeness of extraction; (4) the influence of ammonium sulphate, as specified; (5) the effects of varying amounts of ammonium sulphate; and (6) different methods of use of am- monium sulphate. In the test of the accuracy of the magnesia mixture method, determina- tions were made, in triplicate, on blood, brain, flesh, and liver, with and without the addition of known amounts of inorganie phosphate. The detailed directions followed in this test are on pages 562 to 565, and the results are set forth in Table 1. The results of the further analytical proving of the details of the method were all made on blood. The data are to be found in Tables 2 to 5. PREPARATION OF COLD WATER EXTRACT OF MUSCLE. A. COLD WATER EXTRACT OF MUSCLE. Weigh out 10 to 12 grams of fresh muscle, and divide as nearly equally as possible between two small beakers. Moisten the samples with a few cubic centimeters of distilled water, and break up lumps with a glass rod. Add 50 ce. of water to each beaker and stir contents for 15 minutes. Allow insoluble residue to settle for 3 to 5 minutes; then decant the liquid from each beaker through filters into beakers; allow to drain and add 25 ce. of distilled water. Stir for 7 to 8 minutes, and after allowing to settle, decant onto the same filter. Continue this treatment, using each time 25 cc. of water, until the filtrates measure about 230 cc. each. Allow the filters to drain completely between extractions. Whenever the major portion of the resi- due has become mechanically transferred to the filter, return it to the beaker, using great care not to break the filter paper. After the last extraction throw the entire 1916] FORBES: INORGANIC PHOSPHORUS 563 . contents of each beaker onto the filter, and, when drained, wash twice with small quantities of distilled water. Combine the two extracts, and use for the precipi- tation of the phosphates as described on page 565. B. COLD WATER EXTRACT OF MUSCLE PLUS PHOSPHATE. Weigh out 10 to 12 grams of flesh, and divide as nearly equally as possible between two small beakers; work up with a few cubic centimeters of distilled water; add 25 cc. of aqueous solution of disodium phosphate equivalent to about 40 mg. of magnesium ‘pyrophosphate dividing as nearly equally as possible between the two beakers and proceed as directed under a. The extract thus obtained is ready for precipitation as described on page 565. ; PREPARATION OF HOT WATER AMMONIUM SULPHATE EXTRACT OF BLOOD. A. HOT WATER AMMONIUM SULPHATE EXTRACT OF BLOOD. Weigh out 30 to 35 grams of fresh blood (entire portions as caught from the animal) into a porcelain mortar. Grind and transfer to a 400 cc. beaker with hot distilled water, make up to about 150 ce. with boiling distilled water, place over a flame, and gradually bring to boiling, with constant stirring; when boiling begins add 20 ce. of 20 per cent ammonium sulphate solution, boil, with constant stirring, for about 10 minutes, decant onto an 18 em. filter paper, receiving the filtrate in an 800 ce. beaker. When the liquid is through, lift the coagulum from the paper, being very careful to not break the paper filter, and transfer it, along with that remaining in the beaker, to the mortar. Grind to a smooth paste and transfer from mortar to beaker with boiling 3} per cent ammonium sulphate. Make up to about 50 cc. with the same, stir for 8 minutes, and pour contents again onto the filter paper. After the extract is through, return the coagulum to the mortar and grind a second time, transferring to the beaker as before with boiling 3} per cent ammonium sulphate. Repeat this process of 8-minute extractions of the coagulum in 3} per cent ammonium sulphate, and filtration as just directed, without further grinding, until the filtrate measures about 450 ce. Wash out each beaker twice with 8 to 10 cc. of hot 34 per cent am- monium sulphate, completing the transfer of the coagulum and extract to the filter paper. Wash the coagulum on the paper twice with boiling 33 per cent ammonium sulphate from awash bottle. At all times allow the filter to drain well between addi- tions of extract or wash solution. This extract of about 500 ce. is ready for preeipi- tation as described on page 565. B. HOT WATER AMMONIUM SULPHATE EXTRACT OF BLOOD PLUS PHOSPHATE. Weigh out similar quantities of blood, grind in a mortar, and transfer to a beaker as specifiedin a. Add 25 cc. of an aqueous solution of disodium phosphate equiva- lent to about 40 mg. of magnesium pyrophosphate and proceed as directed under a. The extract of about 500 ec. is ready for precipitation as directed on page 565. PREPARATION OF HOT WATER AMMONIUM SULPHATE EXTRACT OF LIVER. A. HOT WATER AMMONIUM SULPHATE EXTRACT OF LIVER. Weigh by difference from closed weighing bottles 15 to 20 gram portions of finely ground liver into 400 cc. beakers. Add a few eubic centimeters of cold distilled water, and beat up with a stirring rod to separate the particles of tissue. Addenough 564 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 4 boiling distilled water to make the volume up to 150 ce. ; place over a flame and bring to boiling. Add 10 ce. of 20 per cent ammonium sulphate and continue to boil for 10 minutes. Remove from the flame, allow to settle for a moment and decant the boiling-hot liquid onto 18 cm. paper filters. Add 50 cc. of boiling water and stir for S minutes, without further heating over a flame, and decant onto the filter again. Repeat this addition of 50 ce. of hot distilled water, stirring, and decanting eight times, returning the coagulum to the beaker as soon as any considerable amount collects on the filter. With the eighth portion of water throw the entire contents of the beaker onto the filter and wash twice with hot water from a wash bottle. At all times allow the filter to drain well between additions of extract or wash water. This extract of about 600 cc. is now ready for precipitation as described on page 565. 8. HOT WATER AMMONIUM SULPHATE EXTRACT OF LIVER PLUS PHOSPHATE. Weigh out portions of liver as specified above. Work up with a few cubic centi- meters of cold distilled water; add 25 ec. of an aqueous solution of disodium phos- phate, equivalent to about 40 mg. of magnesium pyrophosphate and proceed as directed under a. The extract of about 600 cc. is ready for precipitation as directed on page 565. PREPARATION OF HOT WATER AMMONIUM SULPHATE EXTRACT OF BRAIN. A. EXTRACTION OF BRAIN. Weigh out about 10 grams of brain into a 250 cc. beaker. Adda few cubic centi- meters of distilled water, and work up the brain and water with a glass rod. Make up to about 100 ce. with boiling water; place over a flame, and gradually bring to boiling, with constant stirring. While boiling vigorously (not before) add 20 ce. of 20 per cent ammonium sulphate solution; boil gently for about 10 minutes; allow to settle for a moment, and decant liquid slowly onto a filter of sand on linen, receiv- ing the extract in an 800 cc. beaker. Add to the beaker containing the coagulum 50 ec. of a 34 per cent ammonium sulphate solution, stir for 1 minute, keeping over flame and at the boiling point; decant the liquid onto the filter. Repeat this process of one-minute extractions of the coagulum in 3} per cent ammonium sulphate solu- tion, and filtration as just directed, until the filtrate measures about 450 cc. Wash out the beaker twice with 8 to 10 cc. of hot 3} per cent ammonium sulphate solution, completing the transfer of the coagulum and extract to the sand. Wash the coagu- lum twice with the above wash solution from a wash bottle. At all times allow the filter to drain well between additions of extract or wash solution. This extract of about 500 ce. is ready for precipitation as directed on page 565. B. EXTRACTION OF BRAIN PLUS PHOSPHATE. Weizghout about 10 grams of brain; work up with a few cubic centimeters of distilled water, add 25 cc. of an aqueous solution of disodium phosphate equivalent to about 40 mg. of magnesium pyrophosphate and proceed as directed under a. The extract thus obtained is ready for precipitation as described on page 565. PRECAUTIONS. In making extracts of brain if is desirable that the analyst give careful attention to the handling of the sample. The coagulum is very soft. It should be stirred only enough to keep it in motion. If roughly handled in returning from the sand 1916] FORBES: INORGANIC PHOSPHORUS 565 filter to the beaker it becomes too much broken up and holds onto a great deal of liquid. To prevent the extract or the coagulum from coming into contact with the linen before passing through the sand, pour the extract slowly into a slight depres- sion in the center of the sand, or, better yet, onto a thin film of absorbent cotton 3 inches in diameter, laid over a depression in the sand. The coagulum remains on the cotton, and its return to the beaker is thereby much facilitated. Ifthe cotton is not broken up by needless stirring it can be taken out of the beaker with a glass rod and returned to the sand each time a partial extract is to be filtered. Care is necessary to prevent loss through bumping, on account of sand in the beakers dur- ing the last extractions. Three determinations at a time are enough for one man to handle, but, with some risk of loss, one can handle six. Each partial extract should be boiling hot at the time filtration begins. MAGNESIA MIXTURE METHOD FOR THE DETERMINATION OF INORGANIC PHOSPHORUS IN EXTRACTS OF ANIMAL TISSUES. Treat three of the extracts prepared according to the directions on the preceding pages according to Section A and three of those prepared according to Section B as follows: Add 50 cc. of magnesia mixture, stirring freely; allow to stand 15 minutes; add 25 ce. of ammonia (specific gravity 0.90); cover, and allow to stand 3 days. On the morning of the third day filter, and wash the precipitate with 2.5 per cent ammonia water. Dissolve the precipitate on the filter paper and that remaining in the beaker in which the precipitation was made with dilute nitric acid (1:1) and hot water, receiving the solution in 400 cc. beakers. Neutralize the nitric acid with ammonia; make slightly acid with nitric acid. Add 5 grams of ammonium nitrate, and pre- cipitate in the usual way with molybdate solution. Continue in the usual way for the gravimetrie estimation of phosphorus as the pyrophosphate. RESULTS ON ANIMAL SUBSTANCES. The data in Table 1 show that, as tested by the recovery of added phosphates, the magnesia mixture method gives results apparently char- acterized by a high degree of accuracy. The recovery of added phos- phates was 96 per cent with liver, 97 per cent with flesh, 99 per cent with brain and 100 per cent with blood. In consideration of the close agreement of triplicates, the high percent- age of recovery of added phosphates, and the amounts of coagulum from which the phosphates were recovered, these results are considered a satis- factory demonstration of the reliability of the method. In the further scrutiny and analysis of the method, however, it was deemed advisable to test individually certain of its details. Blood was selected for this work, since the ready decomposition of its phosphocarnic acid was considered likely to reveal improper procedure. The results of these studies on blood are set forth in Tables 2 to 5. Table 2 gives re- sults from a study of the effects of heat and ammonium sulphate in this estimation on a cold water extract of steer blood. This extract was ob- tained through the use of a centrifuge. SAMPLE AND DETERMINATION TABLE 1. Test of magnesia mixture method for inorganic phosphorus in animal tissues by recovery of added phosphates (Analyses by F. M. Beegle). {A = Without phosphates; B = With added phosphates.] WEIGHT OF WEIGHT OF | MAGNESIUM SAMPLE PYROPHOS- |PHOSPHORUS PHOSPHORUS inorcanic | “ADDED AS PYROPHOS- ADDED PHOSPHORUB MAGNESIUM |RECOVERED AS MAGNESIOM PYROPHOSPHATE PHATE? PHATE grams grams per cent grams qrams per cent Blood: Ate TE ran airy ctay 31.30 0.0069 OLO0GE 4S) oe cece ee sd Sa sds,c wenden eee IA RAED 30.00 0.0069 OLOOGAT [1c ceeleull eben eae eee IAS era Gn onerae 25.00 0.0051 OLO00568) (5 5. nn nal ceeeaeee Reee ISIE epadollsonosoaccullsceoosaese 000607 | 2S. eee eee 1335) agerannacc 26.10 OZO558;. iliekieemsicton|its shies tee 020501" ||. ee eee BeQuets eee 28.20 OROSES ail see cers | es-meaket 00501 ae Besii i secon 30.50 OROS U2 Ua Saree Gece poees 0.0505). eee AVOCA gelecs | Sone el coer seen 0.0497 0.0502 101.00 AW 1e einen ces 33.70 0.0064 0400529!) a. actly hs oo | CU ae ei 33.60 0.0060 O:00505 22 ee Arline. dee ee AmB te afeicts sto che 31.20 0.0060 (O200544 9) ox Ferd) nt oe Average: «.::3||Jcueme: 2 |(poneemonns 00526) Soa... Soe Osta eee | See 1 tol ea eeicio aie 30.40 (CHEW Nese eeoseealesanccaper 0.0493 IBa2 astra 32.20 CONE Aca ete mes lis aha Sane 070496 |) eee Loe ow sasBeae 35.80 OLODGSE alee eal beret oe 005001) | eeeneeeene (AVeTape:. 202] 2. a -taaee o| donee ie oem 0.0496 0.0496 100.00 Brain: IAs ee cvente 4 8.5600 Thosty |e Ag tetera seloete eres IIS) cc Aeon. Deas: 7.7011 | 0.0176 | 0.0636 |.......... ies {AS ean acee ae 9.2368 | 0.0209 0.0630 Ce Aa 6 See Averages.sc. [spose eee Q80633" | RSA: 5) See Be Beals scascaes oe LOEZZUS NW MOROSOT. Ni yess eos lsemer en > 020263" 4|-2 eee 12 a sea QE22 77a MNOLOATE® hci secre liiace sa sates 0.0264) |-5 See 12S Cap pons cea LOLSIS 2 MO O502 0a Sonim alin Se peisie =| 020263 10 Beene PA SVETA RG ose a hte inerea tl ee nD, | Noe weep ev pene 0 0266 0.0263 98.87 Flesh: A-1. 13.4653 0.0272 Q10562> eric... calaee 32 cee eee J NDI Oeioexd 10.3444 | 0.0207 0:0557° kee). Oe eee Ans 11517600] te0l0923 |) (00555 mi een n. a lee PAVER ALE one teria TSE anes OFO55S 9 | RST | ee ee Balaastes cain 10896385 0206941125 ., p eee 0.0474 2|\ 2 eae IB=2taahee ease 11.7942 QRO725) lcs eeis ees otha fae 0.0488: --/\8 . SSeeeee 13S ak ea 11.3154 | 0.0708 speaswecwe| WO oe VAWORS Be. Sethe orga seats | ae peters bance hiapeeeeee 0.0496 0.0481 96.97 Liver: H UCIT Cet ne 16.2155 | 0.0627 OV LOR re coc. eae eee ee AED eRe t crs 13.8000 | 0.0552 O.T144e eo ee a ee ARS Owed ee 14.4309 | 0.0519 O00 Riek ote | ese | eee AV CLR PE sy cial| Cranston | pees OFLOG4 WISE: oo AE ee eee 1335 Oe eres 14390941) TOTO45) lo Soetoealee ne eee 050475. |). 20. Baeese B-2.. 14.9658 | 0.1054 rer 0.0482 :|), Seamer Basie ace 16.2232 | 0.1094 penen 10504745 eee B.Qids) yi) (onpeen DR aenee dl tion, 2 eee OE le 0.0496 0.0477 96.16 Ali blanks deducted. [1916] FORBES: INORGANIC PHOSPHORUS 567 In sets A and © the phosphates were precipitated direct, with magnesia mixture, with and without ammonium sulphate added (in the cold) before precipitation. The results were practically identical, and show that, in the cold, ammonium sulphate does not affect inorganic phosphate deter- mination in blood. Sets B and D were boiled with different amounts of ammonium sulphate added. TABLE 2. Effects of boiling and varying amounts of ammonium sulphate in the estimation of duonpante phosphorus in steer blood by the magnesia mixture method. (Analyses by F. M. Beegle). {Cold water extracts. ] WEIGHT OF rata AES ioy || WER GE ee |) oe PHATE F ce gram mg Extract precipitated direct with| A-1........ 300 0.0091 2.535 magnesia mixture DCO eck eee 300 0.0087 2.424 AmB! aiereraversts 300 0.0087 2.424 AV CYA DE Nalin uercpsie revues call oer aris leis: sres 2.461 Extract brought to boiling; am- |} B-1........ 300 0.0079 2.201 monium sulphate added to | B-2........ 300 0.0078 2.173 make 1.25 per cent solution, | B-3........ 300 0.0081 2.257 then boiled for 10 minutes VATE hm telly capeectocin at] (aro emnoeoce 2.210 Same as A, with ammonium sul- | C-1........ 300 0.0085 2.368 phate added before precipita- | C-2........ 300 0.0085 2.368 tion to make 1.25 per cent | C-3........ 300 0.0086 2.396 solution Average’. :il2,srrs starsecs:s/s1e'||>\-seke/creeterse 2.374 Same as B, with ammonium sul- |} D-1........ 300 0.0075 2.090 phate added to make 3} per | D-2........ 300 0.0074 2.062 cent solution DB iiceciatets 300 0.0073 2.034 ARV ETA ROY mellla stole) creccielele evel listlet-fevetavers/onie 2.062 1 All of the extracts and filtrates were precipitated by adding 50 cc. of magnesia mixture to the cool solu- tion, and then, after standing a short period, 25 ce. of ammonia (specific gracity, 0.90). The boiling and precipitation of inorganic phosphates in a 1.25 per cent solution of ammonium sulphate (20 ce. of 20 per cent ammonium sulphate as specified) gave weights of magnesium pyrophosphate 0.15 mg. greater than those obtained from boiling and precipitation in a 3.33 per cent solution of ammonium sulphate. The results were, in both B and D, appreciably lower than those obtained from A and C, with and without ammonium sulphate, but without boiling. These results show, therefore, that ammonium sulphate, in the cold, is without influence on 568 TABLE 3. ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [Vol. I, No. 4 Completeness of extraction and effects of boiling and ammonium sulphate in the esti- mation of inorganic phosphorus in calf blood by the magnesia mixture method (Analy- ses by F. M. Beegle). {Cold water extracts.] TREATMENT! Extract precipitated direct with mag- nesia mixture Same as A-1, A-2, A-3 plus 25 ec. of phos- phate solution Extract brought to boiling, 20 ec. of 20 per cent ammon- lum sulphate ad- ded, and boiled for 10 minutes, filtered, and washed by de- cantation Same as B-1, B-2, B-3 but coagulum ground with fine sand for more com- plete extraction and washing Same as B-1, B-2, B-3 plus 25 ce. of phos- phate solution Same as B-4, B-5, B-6 plus 25 ec. of phos- phate solution WEIGHT OF | | WEIGHT OF i “ MAGNESIUM saurue xo. |YOLUMEOF MAGNESIEM|fymornos:| APPED EHOBTEONDS vaesew ADDED ce. grams gram gram per cent AS eae 300 CON) Kt) dal repens aries Al Sc c)a5 ~ - A-2.. 300 QNOLOS | cis 5 ns en Fee | eee ALS Sete eee 300 OLOLO3 i soe8 aces bere cee See eee Average OL OLOS | «cece :< aczselfles ane one oil it Aad! ctee 300 020619 oo cas dalle oo ee eee eR ee A-5.. 300 Q20618 jee hoot cone eee A-6o0.6G88 300 00620) ..00.24.:2 Jalcce coc ee lhee eee Averagess. issn. cece = 0 0619 | 0.0517 | 0.0514 99.42 B=15-25-254\— 1300 020083))| : 3. ete eee Bag ase 300 (A Uoio i eeeeete ein tera laicgach oe: B-3t.5..--5|) 300 00091 | ssc ss es gee eee AVCLARE ss. Aa ctr) (OsO0S8T alan. oe Bin Og< i) moo ze a ae s& | PHOSPHORUS = * 5 mo f=) O26 | RECOVERED AS SUBSTANCE TREATMENT 2 6a z EEE aii maniera 2 Bx Sy Fs = 6 |PYROPHOSPHATE = obs a anak a ADA iS} B<8 2 = 5 < gram |per cent) gram gram { | Filtrate precipitated with 50 1 | ec. of magnesia mixture plus 2 25 ec. of ammonium hydrox- 3 id plus paper pulp Blue grass......... } Average |; Same plus 25 ce. phosphate 4 ) solution 5 | 6 : Average {| Same as 1, 2, and 3 ;| Same as 1, 2, and 3, plus 25 cc. 10 0.0225 Rice polish......... 4 phosphate solution ie : Averagel|.2. 226 |s socsceeleceanne “0.0225 i Filtrate precipitated with 50 1 ec. of magnesia mixture plus 2 | 25 ec. of ammonium hy- 3 droxid Average |........ 010424 |... 00]. See eee | | agsnqoTonsaasecaosodsor ase 5 4 0.0543 |........| 0.0768 | 0.0467 Rice polish plus 5 010461) eae e 0.0768 | 0.0385 phosphate....... 6 0.0443 |........] 0.0768 | 0.0367 L . Average] Maricc-nis|a-a nee arene oer eee Fanacgonsawecoonssssdsade Cane 7 Middlings.......... { . | Average | O30131) }0[07245| Saas | eee eee es 1916] FORBES: INORGANIC PHOSPHORUS TABLE 7—Continued. SUBSTANCE Middlings plus phosphate Soy beans plus phosphate....... Soy beans plus phosphate....... Middlings.......... Middlings plus phosphate....... Oat straw.......... Oat straw plus phosphate....... TREATMENT Usual method, plus filter pa- per pulp; without phosphate Usual method, plus filter pa- per pulp; with phosphate Same as 1, 2, and 3 of this set | Same as 4, 5, and 6 of this set | Same as 1, 2, and 3 of this set Same as 4, 5, and 6 of this set SAMPLE NO. PYROPHOS- WEIGHT OF MAGNE- SIUM PHATE a Ba Bee cag 5 a & a PHOSPHORUS Dd 6 <3 | RECOVERED AS ta Dad MAGNESIUM y — © |PYROPHOSPHATR a aa a g B...-t-ekeees iene reece epee ere 411 Casein, precipitation, paper by Van Slyke and Winter........................ 281 Cereal products, recommendations by Committee C........................25. 286 recommendations by; Whitey.c.- 22. - « sa-4- eae cies 199 report, by | White ss. 26)... s%, los eee 337 INDEX TO VOLUME I 593 Himeryerepord) Onysymuheticnproductsn- emma ase yrte setae eetceireieiclcrc ele 337 and Palkin, paper on estimation of antipyrin, reference............... 343 Emmett, report on separation of nitrogenous bodies (meat proteins).......... 267 Bixner paper on thersublimabonic. 222 c ts ane eens cerecinains icici ov 208 Extract, ginger, paper by Harrison and Sullivan................:...-....20055 506 Extracts, flavoring, recommendations by Paul..........................5- 153, 505 PEporteby hauls sy Wak eects cee eee ero olerele eee 146, 498 Fats and oils, recommendations by Committee C.....................0 00000 ee 285 recommend awonsi bw Weli weer rms cee a eee ae 186, 515 TE POTD NOLL cea <= opetess ods ace ale ois res ore de eveaeteheene e 181, 513 Feeds and feeding stuffs, adulteration, appointment of new associate referee.. 107 recommendations by Committee B.................. 331 recommendations Dy; JONES +. hee ase ee eee 312 MEPOLES DY; DOMES eam ane ets cet cine yess Mee eee: 289 Feldspathie fertilizer, potash content, paper by Miller and Vanatta........... 26 Flavoring extracts, recommendations by Committee C........................ 284 recommendationsibys baulunasases sceieceeeeeee eee LOOM OUD TOPOL, Diy aul te rreeetepctete creneysje/steregs cate ois etersieks oie oi ste ea ere 146, 498 Hood wadulterationy report by, Elortvetse--eeersee cease casein deuce aces. 110, 465 definitions, report by codperative committee (Frear).................... 462 heavy metals content, recommendations by Committee C............... 287 recommendations by Loomis..................... 254 recommendations by Treuthardt................. 589 REDORUW DY, WOOMIS s.r ei tae oes erate ee see 244 MeV Lon? ARAVA cespoeopedooneoneHopooueaee 580 supplementary report by Treuthardt............. 254 phosphorus content, recommendations by Committee B................. 333 report by; Horbes) and) WussOw. 225: 420+ seas: 221 standards, report by committee (Mrear)............-..-+-.s+++++s-- 108, 461 mnEcontenteLeponi py Mireuthandterceenceie fice cee eeaee eenee 254 water content, recommendations by Committee B...................... 332 mecommendstionsibys Vic Geexaa.cssscieeeecis en nance eae 221 TEPOLU DY » McGee teases ciisie yeasts a apace oeusieroetra seine 218 Forbes and Beegle, report on inorganic phosphorus in animal and vegetable SUDStAMC eS pee erate yee versace ovasa/cishc tba tes Poval Ski atin aise: « 562 and Wussow, report on inorganic phosphorus in foods................. 221 Fraps, paper on interpretation of soil analyses......... ..........00+.2++--- 418 POD Te CERT ENG CGE), ob CREO SIRS AG 6 On SERRE ao aR RE, on 158 TEPONt ONS l SeeeR Merce ieee ee eh ea tet ne ee iran eta saisear 33 Frear, report by committee on food standards..................-eeeeeeess 108, 461 report by codperative committee on food definitions.................... 462 Hrurt jellies; papenibyeAmdnews, reference: ....4.. 02.00... 40-cneocs- cesses 130 JWicess papersbyeAmonewsrreterencessten yan-sn cae eens ones ce se eeaeees 130 products, recommendations by Committee C..............0.00 ccc eee eee 282 NECOMMendabions) bya Gorevsesss- ns aces ket eee casa cre 130, 485 LE POR VEG ONE wee eT neat et ee ee eA os lee 120, 480 Ginger extract, paper by Harrison and Sullivan......................e--0ee0+s 506 Glycerin, in meat extracts, paper by Cook 279 594 INDEX TO VOLUME I Goodnowastreport (OM VINE LATS 26. 6:s.2-.- sed ow se eet aes een ee ee eee 145, 496 Goreysreportion fruitiproductsss. ............2¢.000.++s-5s00s ss eeene 75 recommendations by Committee A... ... 5.2... o0s e200 ees see ee 101 TECOMMEeNnGaALlons|DyehOALken eee me eine< = ocean Eee eee 456 TEPOLt Dy AVELLb bre rectors srs c eleic eres jeie la bid seve 6 5/5 cise ere aces eae 59 TEpoOrtrby oar kere eats ccs style eo sarsereie seca le secre 435 Jarrell paper on deternminationofspotashis-s- ssh ce aes. oc - see e eee 29 report on determinationvorpotashe se. ce sees c sess ee eee eee 400 Jones, paper on lime requirement of soils. ....:. 2.2.2... 0... cece c cs oer ees onee 43 reporsonteeds andweedingystuliss.a.-. a seeeneceee cack aa eee 289 Journal of Agricultural Chemistry, report by secretary......................-- 523 CISCUSSIOD.. «teen «+s - Bee oR Cee 531 Journal of Agricultural Research, endorsement........................20-+--0 335 iKerrsreport.oni fats andioila yan see ose eee eter 27" os eee eee 181, 513 Whadd; president/s address vss. qtaceitect hisses ne Seo anc ao ee ee 515 Lead arsenate, water-soluble arsenic content, report by Averitt............... 74 Lead, in baking powders, paper by Seeker and Clayton......................- 264 Leather waste, nitrogen content, paper by Rose.....................--2-2200- 396 Lime, neutralizer in dairy products, paper by Wichmann, reference........... 195 requirement/of soils, note:by, Veutch). ...n1-.-cee sos.) 2 eee eee 44 paper by, Blairiand Meliean:.2..4..).-2 eee eee 39 paperby Jones cxha5 \G. teedssen ana eC eee Cee 43 paperiby McIntire co. ccccne neces eee eee 417 recommendations by Amés..................--«-<: 416 INDEX TO VOLUME I 595 Time-sulphur solutions, paper by Averitt. ..............eccee sees ret r steer ens 95 paperdby Ro arkeaner mets sae serie cisvisthhevas x serevcus Sai 76 recommendations) by; Averitt... .4..0-5++2c2++-- 00 75 recommendations by Committee A................... 101 recommendations) by; Roark. ssasenes sane sey sei =e «2 93 : Teporbi by Awertbtinss varyeccrecurcttetenmebse eaters erly ay steer falc oon 60 Lipman, report on nitrogenous compounds in soils..................-+- +020 eee 422 WsGomrxesreportiony heavyrmetals) tn TOOdS seme mea on ce selec eras ele ele seleieie «oro ss 244 Lythgoe, report on cocoa and cocoa products.......2.......2..eseeees ess 200, 550 McDonnell, report on determination of potash...................se0cceeceeees 22 NeGeeyreport.Oni water in) TOOGSe i ..)os atmciete a sis) s sina sone) cle, src.nciclelsherbemiat eet ekercye 218 Mcintirespaper on) limemequirement! of soils... sus. 2. «> ates a siie ene aeons 417 and Curry, report on inorganic plant constituents.................. 55 and Hardy, paper on effect of ammonium carbonate on determina- GLO O fp MUTTUS epee apsietes ooh sista cseictrclactere sites atayeisys. satel suchas siapesieerave cleisia ears 44 McLean and Blair, paper on lime requirement of soils......................055 39 MUO Er Te POLtLON Ve LOtADIES 2 7-.nte10-o rela asi ocas ox isle erere ite ieie sas vas Sesto efafei soe 199, 545 MrraAschinompaper by, hiley, andi Sulllity ames sieerci-tersl-is ors iatcexcereilerncvenreisiereerelors 490 Marti and Patten, paper on method for preparing neutral ammonium citrate SOUT, WAR Sat hon ne apd Soon racpaodnno aco nccac mapa nis tecors Dake cemartce 17 Wianhewson report: on cOlorsiss. eet: chascrian rae eitios eeeasitesio eee 113, 470 Meat and fish, recommendations by Committee C.....................2220205 285 MECOMMEeNnd ypONS| ys Mb eee eel sieeeieleielseeraelel ale 180 TOPOL Vgrowen Ut Ley crets cere rays seer or crass ope ere ee eee ere 186, 515 report by Kern! << 5..92$etc paper by-Exnenasceee ern saint ees ieiieels eee = eee ee eee 208 Sugar and molasses, recommendations by Committee B....................... 332 recommendations by/Crossemsecise | cscieteie eie)iel-leiei ole eee 317 TEports bys Crossssccscceeyoctesaerie eres she oho tee eer ee 314 Sullivan and Harrison, paper on ginger extract...................--.22e+-0e0s 506 Riley, paper OnymMaraschino se. o-ycciscctee tes ere ell eae) lice ee 490 Synthetic) products, teportipy, Pimeny.. oii cee teenie eile eee 337 Tannin; LEPOre, D¥7 BACOWs ca. meee ec leleiee « Scierisice clon arora ele ieee ater rere 329 Tea and coffee, recommendations by Bartlett.....................0.0se0es 208, 556 recommendations by Committee'©................---.--ye--== 287 TepOLbibysbanulebuenrerete eect ees ee eee 203, 552 Tin, in foods, recommendations by Treuthardt....................2.......005- 589 TEPOLbIDyadce Ubh are teeter ele eyed ero ster cte ister tee 254, 580 Treuthardt, report on heavy metals in foods.....................+20+002-e00es 580 supplementary report on heavy metals in foods: tin.............. 254 *Eriammonium (citrate; Papen ysl allece ee ccinctereelsctei tenia © creeete elects) atelier 375 Vanatta, report on availability, of potash=. .- ...eeecers- 2. ee eee 24, 398 and Miller, paper on potash in feldspathic fertilizer................. 26 Van Slyke, report by committee on study of vegetable proteins............... 109 and Winter, paper on precipitation of casein....................-- 281 Vegetable proteins, study, appointment of committee......................-.. 157 noteiby, Alsbergis.assecrecs o. 5 cccece esate eee 464 report by committee (Osborne).................... 462 report by committee (Van Slyke).................. 109 Vegetables, recommendation by Committee C.....................200 20022 eee 286 recommendationioy Mapruden: . ne ceial-a towtey rrspsn oa iomeah seo pve ony Sue Ih HW meaterrana peter yy mere oareraenee ree Poe pense powers earn weet pene toate oOo: rh peer wenger ey Kocntibpantesneptn ses jones tno aeeaneal rece vod tented rte $00 Bahan o- Ooms dekep enh et ie poperoresreee yeeeraee taunt $ army rn rcp fant am rane va ep lw orieernscatitalasg thstater ea te : serene sere nately mite eiearnegraeee sieh pcb cate eee ca Nosteneysseose acest &) reneerre OS aa per ase stata Saat Hit prerseee tvyrb se ityeyretire sa titer wehterier aa ener cargoes me bat sent am tel Vinee ae teahabe geben beentnd 06 90-44 1066 eet esrorenneeveetea- 9 36% thoy ¥ ot Coltiens berry norton, atch AetP RN OA 8 waters te D9ebiiga he pd aceutek fad read cent ot Aeneas a Sseer wrenney ) ¥ ier vet 4d hahha ee PoE Te areal tat ged shone at cS jabs Te pat nae hoe teh ese ay teen caaeag vet sahara be per be bd me Creriper beter ety iets Si hen aergseera rieas anes beh . COI pclecind hese wene rural pebeyeserey Star) wrote reer EO ety a apreew er ertrert tert et cp rhateaeten bv beam By igtinde ber tae Ort puteatinde fret etal rei be os werierenny Toke hat yt pebenatnte Corer sarees . forse ech Preah ceolbessens ae sterh fetetet Seek Net eicbehd aber weceetedey Oe borate ied ob se tasonenonest 18 O24 grtsirgl a: gat Weahe tate iced otet fabnpe tase ett steel ard a bab what ve for POC eee sgrareretr tate! t are ede aprcees see tine seen 10nd fw, wt ebeceocineng sites sentamrenecsaton yf cevesenen eet Npiit besos tensesantase fhe ; rere) see pero} vet eT toe vetinaet abana i ogous 86 erere erento erewraeyet ors rs sated gaactite ‘i ee] = Savon oye eh Orda. ber rn abe Neat peeve stot) pores messereenmenmrsc uy YOST he pat ninte tie t + aot oo eeebcander Sreeatatettis pierert ioe beeen btdegubaes stave — cbsin'sbeieaehegegantinet st ob -oeFet mah pe NSS EO see sitet ; esa ore eT of doe Otnetabe wes 1 eaisineceeba bth Va Tnabeens poe Cherie Wk bot sateen sine ae re a ether Sevens wicetsseaveeeet errors ro ‘ pe ee aan’ jaicka tego abet svines a> (ress oor peue prints tneahatpn sorters ree ehh pecrewies ne Arpt e ag o eityetatninda te ht3 toe shardsinmys acct BtcOrsetal sett tee rayfeneiee ee eS oaenetrrree corarieacetetaey treet Fo: ii nea aek teers ee ness Cha paneer neewy seit renerries rire 1 eed Cy epee nist tereehan bess ise ened ques his oso OeTienhcs pears eee teteend cates SSoeagyrenasiee see rsd ied cory ecses speanbooteactstesseresas st egceress Or eset 4 ste reper n doe r eich oa pote est botee tone ntme nd to neonnd re sdarorent sont Seasiieapt webb Sitter ie ott RE cE EN RE Re oot ee a ie tent a eater ses ais dutch oeed ated iad aeerend novindn dnd Anendntithtoret og 1OrmGn ald Ge BY Gbenties® fotared enh rene teerdod-' eet oe Mectnaetee eit: Sci itewers net Sih hres cerererore ns oo atere renames Ligh 4 beh eat Te eee Pannen ar et anes fore Seas ep entint anh "eatery teeta eet ae ne cenane ee Cares mt Teehd a de Ooe 4 ras @rerbsChanaee +: 0 Seed satan eho i He Ob rate nantns ne piginey deme tnt # speerevesed otbeeitoredes Poets nveree as ” ar babe be roleh