ACCURACY IN THE ANALYSIS OF DAIRY PRODUCTS. - Journal of

H. Droop Richmond. J. Am. Chem. Soc. , 1893, 15 (10), pp 579–591. DOI: 10.1021/ja02120a009. Publication Date: October 1893. ACS Legacy Archive...
0 downloads 0 Views 668KB Size
ACCURACY I N ANALYSIS OF DAIRY PRODUCTS.

579

tempted by the writer of this communication; the results form a volume of over I , 2 0 0 pages just issued by the Smithsonian Institution as one of the series entitled Miscellaneous Collections. This " Select Bibliography of Chemistry, 1492-1892 " embraces about 12,000 titles in twenty-four languages, yet makes no claim to completeness; it is moreover a bibliography, not an index. I n conclusion I have the honor to propose the appointment of an International Committee on Chemical Bibliography, to consist of one member from each country represented in this Congress. T h a t this committee have unlimited power to add to its number, provided however no country have more than two representatives. That this committee, through the European and American Chemical Societies, report a scheme for an International Co-operative Index to Chemical Literature. UNIVERSITY CLUB, N E W Y O R K CITY,

August, 1893.

[EDITORIAL SOTE.-A committee consisting of Edward Hart. R . B. Warder, and Wni...I Dudley was appointed by the Congress to consider the recommendations of Dr. Bolton set forth in the above paper. Their report with additioual action by the C o n g r e s s will be found on p . 309 of the June number of thejouvnal for this year.]

ACCURACY IN T H E ANALYSIS OF DAIRY PRODUCTS. BY

H. DKOOPRICHYOND, CHEMIST T O T H E AYLESBURY D A I R Y CO.

PART I.

I

MILK ANALYSIS.

N the July number of the Analyst for 1889 I published a paper on '' Fat Extraction from Milk Solids ; " the object of

that paper was to show how the differences between the two rival processes of fat-estimation could be explained and rectified ; I succeeded eventually in obtaining identical results by the processes due to Adams, Soxhlet, and Storch. Since that paper appeared it has been showfi by Waller, Liebermann, and others that the processes of fat-estimation, especially the Adams method, should be modified by rendering the use of ether as a solvent inadmissible. T h e Werner-Schmid process has also come into extensive use. I n considering the accuracy of any process we have three 1 Read

before the World's Congress of Chemists, August

22,

1893.

methods of obtaiiiitig evidence oi esactitude : hrst a synthetical iiiethod, that of taking kiio\vri cluaiitities of th,, iiitist:uice to 1)c. estimated, niixiiig it with the s1ii)st:lilccs froiii \vliicli ir is t u l c separated, aiitl perioi iiiiiig a11 c1. i i i i : i : i o i i : i f tiii- siiiistxiicc ! lvhile. for iristatice. for Holstein co\vs they ivoultl !)e i i i iiiaii>. iiistanccs too liigh. Icveii raking shortlionis the iiiiiits are in rare cases overstepped. I11 t h e milk of siiigle CU\YS 1 ha\-e records of fat as io\\ as 2 . c ) 1)er cent. aiid solitis not fat d o \ ~ ito i 7 . ; per cent. i i i genuine s;iiii~)Ies. It woulti be iiianifestly uniair for a chemist to slvear tliat el-ery smiiple giving 2 . 0 percent. of fat. or ;.;prrceiit. of solids not fat was genuine. nierely lxcatise oiie genuine saiiiple out of iiiaii!. tIio~isaii(isexainiiieci gave sucli figures, arid equally s o to ~ ~ J ! l d ~ i l l e:.er~ll sample that :vas just lielo\\. the limits adoptetl. \Vith regard to low percentages of faL I see i i o \vay at present of disc*riiiiiiiatiiig bet\veeii geiiuint inilk lo\v i n fat, and iiiilk froiii wliicli part of the fnt Iias heen abstractect. I alii not ivithout hope that this niay lie accoinplishetl. Tlie size of globules, aiid the proportion of those of l a r g e to those of sniall size ma). afford n iiieaiis to this end, aiid as creaiii uiitloii1)tecll~- contains a higher proportionate content of albuniiiioid, it i:, possil~le that ail estimation of a deficiency of this i i i the iiiilk ilia). afford ei.itlciice of fat-abstraction. I have no evideiicc to offer on this p i i i L , but put forivard the suggcstioiis for the coiisideratioii of the Congress. \t'itli regard to loiv solids not fat, iiiucli evidence can be obtaiiietl from a fuller exaniiiiation. T h e whole of m y experience goes to show that geiiuirie niiIks low i n solids not fat never furnish an ash of less than 0.7 per cent., while the majority of watered milks give a n ash below that figure. Further the excess of ash iii genuine niilks is not due to chlorides or ail). watersoluble substance, but to insoluble phosphates. The ash of inilk is neutral, is entirely soluble in dilute acids, and contains no boric nor carbonic acids. 1 Iiai-e not found the nitrogen i n genuine niilks to fall below 0.j per cent. X curious coincidence is that the percentages of nitrogen and insoluble ash are nearly the same in niilk. I find an average of 0 .j3I per cent. of nitrogen and 0.530 of insoluble ash, with extreme differences of 0.02 per cent.

A C C U R A C S I N AN.4LSSIS O F D A I R Y PRODUCTS.

587

As many waters contain nitric acid, which I have never found in genuine milk, the diphenylamine test for nitrates is of service in deciding whether a given sample is adulterated. I curdle the milk by heating to near 100' C. and adding a' small quantity of acetic acid, and filter the serum. I place a little diphenylamine in a porcelain basin and add one cc. of pure sulphuric acid, and allow a few drops of the serum to trickle down the sides of the basin over the surface of the acid. If nitric acid be present a blue coloration is developed in ten minutes round the edges. I have detected five per cent. of water in milk by this test. If I find a milk of (say) eight per cent. of solids not fat, and I find in it nitric acid, and the ash falls below 0 . 7 per cent. and the nitrogen below 0 . 5 per cent., I condemn it ; while if there is no evidence of nitric acid, and the ash is above 0 . 7 per cent., and the nitrogen above 0 . 5 per cent., and I find no extraneous body to account for the excess of ash and nitrogen, I decline to certify that it is watered, and should state that there was the strongest probability of its being genuine. I may cite examples :

. . 1.0280 1.0295 ...... 10.63 11.62 .............. 2.72 3.47 .............. 0.62 0.64 .... 7.91 8.15 ....... trace trace

Specific gravity Total solids Fat Ash Solids not fat. Nitric acid Conclusion

......

1.0zg0 11.31 3.22 0.76 8.9

none .watered watered gen.

r.ozgo 11.37 3.25 0.72 8.12 none gen.

1.0285

11.36 3.30 0.72 8.06 none gen.

1.0285 11.99 3.87 0.71 8.12

none gen.

1.02~ 11.34 3.20 0.72 8.14 none gen.

I did not determine the nitric acid in these samples, as other evidence was foithconiing to assist my conclusions. T h e season of the year should also be considered. T h u s in England the recent drought in June and July has caused the niilk to be of poorer quality than has hitherto been recorded. It would be advisable in every country to establish a Governnient Laboratory to study the niensal fluctuations of the coiiiposition of milk, and to communicate the results to chemists charged with the working of the adulteration acts. In Eiigland no such laboratory exists but the function is very partially filled by a private dairy company of whose laboratory I have the honor to be in charge, and 011 numerousoccasions I and my predecessor have been consulted by public analysts. The apparent

588

A C C I ~ K A C SI 9 . ~ S A I . S S I SO F I).LIKY PKODTCTS

disadvantages of a trading coiiipan). under tlie supervisioii of the adulteration act performing a function properly belongshows the necessity of the establislinient of ing to the state, oiil>~ Goveninient Dairy Laboratorics. Other countries are iiiore fortunate than Eriglaiid in this respect, mid I 1iaJ.e no doubt that representatives of those countries will testify to the usefulness of these laboratories. E s fim a tion oy Milk .Srrf=ii r .-Vi et 11' s modi ficat ion of 'vV iley ' s iiiethod of polarization ( A n a ~ , ~/6'88, f , 63 j gives results tvliicli agree well with the gravimetric iiiethod of estimation b y Fehling's solution. I alii iii tlie habit of iiiakiiig a i l estiiiiatiori of the copper oxide reduced by a k~iowiiweight of pure milk sugar, as nearly as possible equal to that in the quantity of iiiilk tnkeii for analysis. perforiiiiiig tlir two estiiiiatioiis uiitler the same condit i oils . T h e probable error of each estiriiation, and of the two estiiriations together was fouiid to be i.0.06 per cent. The ivork of Wiley 0x1 the polarization of milk sugar is so coiiclusive that it is not necessary to search for further proof of accurac!'. but this deterriiiiiatioii inay be accepted as one which caii he cleperidecl upo11. Z