I N A RECENT communication of ours,’ certain quantitative methods were described for the chemical determination of various nitrogen fractions in solutions of allergenic substances. In one of these, trichloracetic acid was used as the precipitant for protein, and a quantitative analysis of the nitrogen in the precipitate was carriecl out. Nitrogen values by this technic were compared with the values obtained in the same solution by precipitation of the protein with phosphotungstic acid. In every instance the nitrogen precipitated by trichloracetic acid was less in amount than that thrown down by phosphotungstic acid. In the discussion it was pointed out that this difference was due to the fact that trichloracetic acid is a more specific precipitant for true protein than is phosphotungstic acid, as the latter carries down other nitrogen-containing substances as well as prot,ein. In our conclusions me suggested that trichloracetic acid might prove of value in t,he chemical standardization of solutions of substances whose allergitally act,ive principle is known to be a true protein. Spain and Newell,* by ultrafiltration through cellophane, removed the allergically active substance from ragweed extracts almost corn. pletely, but reduced the nonprotein nitrogen concentration only slight!y. Several workers39 4 have shown that. protein is removed from allergen solutions also by passage through bacterial filters, while the nonprotein nitrogen decreases but little under this treatment. By repeatedly passing a timothy pollen extract through Seitz filters, we demonstratedl that all nitrogen fractions were decreased although in diflc~rent, proportions. The percentage loss was greater in the nitrogen precipitated by trichloracetic acid than in the nitrogen precipitated by phosphotungst~ic acid, and least in the total nitrogen. The repeatedly filtered extract had undergone only a quantitative change, since there is no evidence that filtration alters such a solution qualitatively. Therefore, a method was at hand for estimating the relative accuracy of each of the nit,rogen fractions as a measure of allergic act,ivity, by comparing the repeatedly filtered extract with the original extract by means of skin tests.