Thirty-l;\-e milliliters of an aqueous solutioil roilt:linii~g 5 to 70 nlicromoles of \.olatilc C! to Cs acids was dislilled to constant volu~llc without reflusing. The tlis~illa~c xvas titrated, made alkaline, and exxporated LO ciryllcss a t 110°C. The free acicls were regenerated by the addition of 0.03 1111. of 10 N sulphi~ric acid.absorbecl O I I lilter paper, and estracted clc~atltitati\.ely ivith 3 ml. ol chloroforml,~~tanol (09: 1). Acids present ill the estract were separated by chromatography or1 a colnposite silica-gel col~~nlll employiilg tivo ir~ter~lal indicators, and after elution were tlcterrninetl by Litration, the tlctails lor which are given. The ~~r e p n r a t i o~l ot a reproducible silica is descril~ed.
Introduction,. 1 he use of silica gel for the quantitative separation of snzall a m o u~l t s of iatty acids, mnging from acetic t o butyric (2, 12, 17). valeric to clecanoic (lSj, and ~~lldecanoic to stearic (10), has been reportecl from several laboratories. In general, different solvents (both deve1ol)ing aiicl stationar!.), separate col~umns, and separate samples of acids have been rccl~lirccl for the analysis of each group of acids. T h e only methocl reported which attelnpts to bridge the gap between two groups of acids is not presented in suf5cient cletail (15). Difficulties have been encountered in the preparation of rcproclucible silicas ( G ) , ancl there is rlo method available for the quantitative tralisier of milligram quantities of thc lon.er f a t t l~ acids from the aqueous s o l u t i o~~s in n.hich the)-usuallj, occur to the s~nall v o l~~m e s of organic solvents r e q~~i r e d lol-t h c chromatographic anall-sis ( 3 ) .Our investigation of the fatty acids excreted b). nematode parasites made the use oi a single small salnple imperative, clespite the fact that these acicls varied in coustitut-ion from acetic to hexanoic and higher. Moyle, Baldnlin, ancl Scarisbrick (11) successfully separated the acids in Lhe acetic to octanoic range by the use of 1-hree separate buffered-silica colunlns, together with a fractional titration of the eluate. Unfortunately, their methocl is laborious and ~lnsuitable for nul-nerous analyses.As already reported (4) it has beell possible to modify the original methods of Ramsey and Patterson (17) and of Elsden (2) in such a way t h a t acetic, propionic, butyric, valeric, hesanoic, and octalloic acids may be quantitatively 1