of these brands of gelatin have a higher jellying power than Gold Label gelatin; so that if they are made up in a strength of 10 per ct. about as firm a jelly is formed as with 15 per ct. Gold Label gelatin.

The results of this work are given in Table III. In Part I of this table, various concentrations of Gold Label gelatin are compared with two different concentrations of U. S. Glue Co. gelatin; in Part II, 15 per ct. and 20 per ct. Gold Label gelatin are omitted, but three strengths of Bacto-gelatin are included. The omission of 20 per ct. Gold Label gelatin in the latter part of the work was simply because it did not allow colonies to become large and hence did not favor qualitative work. From a quantitative standpoint, however, it has made a very good showing, as it gave a higher count than 12 per ct. Gold Label gelatin in three of the eight tests and an equally high count in one of the others. Its use is highly to be recommended if quantitative results alone are desired, as its slow liquefaction makes a long incubation possible even if a temperature as low as 18° C. is not available. Eighteen per ct. Gold Label gelatin was given a more thoro test merely because it was better for qualitative work, as will be discussed later (p. 114).

Of all the different formulae in Table III, the best quantitative results were obtained with 12 per ct. U. S. Glue Co. gelatin. It has given a higher count than Gold Label gelatin in 25 of the 38 times when a comparison has been made between the two media. Its only disadvantage was for qualitative purposes, as it did not allow certain Actinomycetes to show up as distinctly or with as great chromogenesis as on either of the other two brands of gelatin. Ten per ct. U. S. Glue Co. gelatin was tested only six times, but in those six tests, it gave as good results as did the 12 per ct. formula.

Bacto-gelatin has made a rather poorer showing in these tests as it has equalled the count on Gold Label gelatin in only five of the fourteen tests. This apparent inferiority, however, may be due to the fact that this gelatin is so purified at the factory that it does not need to be clarified with white of egg, and the materials contained in the white of egg may serve to increase the development of bacteria on media to which white of egg has been added. In that case, it ought to prove possible to find some substance of definite chemical composition which could be added to Bacto-gelatin in the place of white of egg; and thus its apparent inferiority would prove actually a point of superiority. No such substance has been found as yet, and the tap-water formula has proved the best thus far investigated.

As a result of these tests, the following formula is recommended as a gelatin medium for quantitative purposes:

Tap-water, 1000 c. c.

Gelatin 200 g. of Gold Label gelatin.

or 120 g. of Bacto-gelatin or of U. S. G. Co. gelatin.

Reaction adjusted to 0.5 per ct. normal acid to phenolphthalein. (This will require about 20-30 c. c. normal NaOH for Gold Label gelatin, but under 10 c. c. with either of the other brands.)

Clarified with white of egg in case Gold Label gelatin or U. S. Glue Co. gelatin is used, but unclarified if Bacto-gelatin is used.

As

Comparison between plain asparaginate agar and asparaginate agar with glycerin.-The asparaginate agar (formula given in footnote No. 36), has been modified by the addition of 10 grams of glycerin per liter. This was found to favor the Actinomycetes, to stimulate chromogenesis and to increase the size of the colonies to such an extent that it was immediately recognized as an improvement provided it gave a count as high as the unmodified formula. The 25 tests given in Table IV were made to investigate this point. In these 25 tests, the asparaginate-glycerin formula has given a higher count than the unmodified formula in 12 cases and the unmodified formula has given the higher count in 11 cases. the difference between the two counts has never been great, this seems to indicate that one formula gives as good quantitative results as the other. The two formulae have also been compared in an indirect way, with results that distinctly favor the asparaginateglycerin medium. In Table IX of Technical Bulletin No. 38, 96 comparisons between gelatin and plain asparaginate agar were given. In only 34 of these cases was the count higher on the asparaginate agar than on gelatin. Add to these results the data given in Table IV of the present paper and we have 121 tests in all, in 55 of which the count was higher on the unmodified asparaginate agar than on gelatin. Now in Table V are listed 72 comparisons between gelatin and asparaginate-glycerin agar, to which are to be added the 25 comparisons given in Table IV. In these 87 tests, a higher count has been obtained upon the asparaginate-glycerin formula in 67 cases, a considerably better showing than that made by the unmodified formula. It seems well, therefore, to recommend the glycerin formula, because it gives as good or better counts than the unmodified formula and allows greater differentiation between the colonies of the different kinds of bacteria.

Comparison between gelatin and agar.-Whether to use this asparaginate-glycerin agar or the tap-water gelatin will depend upon the conditions under which the work is being done. The data given in tables IV and V show that the agar gives, if anything, rather higher counts than the gelatin, but does not do so consistently

enough to make its use advisable in all cases. The counts obtained upon the two media are nearly enough alike so that if a medium for quantitative work alone is desired, the choice between them should rest upon other considerations such as whether low temperatures are available for incubation. If, however, qualitative results are desired, each medium has distinct advantages of its own. In this case a number of other factors besides the total count must be taken into consideration. A discussion of them is given in the following pages.