226 Report of the Department of Bacteriology of the 



ority of gelatin so far as count is concerned, as in only two tests 

 (Nos. 8 and 9) was the gelatin count appreciably different from the 

 agar counts. Sometimes one medium has given the highest count, 

 sometimes another. None of them has given consistently lower 

 counts than the asparaginate agar, with the possible exception of 

 Lipman and Brown's medium, which has given a higher count only 

 twice, and has equalled the asparaginate agar in count only three 

 other times. These results, considered together with those listed in 

 Table XI, give a particularly unfavorable showing to Lipman and 

 Brown's agar, a fact which is important when it is remembered that 

 this medium is the only one except the asparaginate agar that does 

 not contain an appreciable amount of any substance of indefinite 

 chemical composition. The differences between the various counts 

 in Table XIII, however, are all too small to be of significance; and it 

 must be concluded that under favorable conditions any one of these 

 media (with the possible exception of Lipman and Brown's) is adapted 

 to the growth of as many soil bacteria as any of the others. In 

 choosing between them, the decision must be based upon other 

 matters rather than upon the number of colonies they allow to 

 develop. 



Important considerations to be taken into account are these: 

 A great drawback of Fischer's agar is that the colonies are all mere 

 pin-points and cannot be distinguished from one another. A serious 

 disadvantage of Lipman and Brown's medium and of Brown's 

 modification of it is that molds and overgrowths are often so abundant 

 upon them as to interfere with the counting and prevent the isolation 

 of pure cultures from the colonies. A further objection to Brown's 

 agar arises from the difficulty of obtaining an even distribution of the 

 albumin, which must be added after the medium has cooled enough 

 not to cause coagulation but before it is cold enough to prevent 

 tubing. Temple's agar proves especially attractive to Bacillus 

 mijcoides, which was so abundant and vigorous in some of the soils 

 studied as to overgrow the plates and to render counting difficult. 

 Considering these points in addition to the advantages of the two 

 new media that have already been discussed — the superiority of 

 gelatin in the matter of allowing distinctions in appearance between 

 different colonies, and of the asparaginate agar in the matter of 

 definite composition — it must be concluded that the gelatin is the 

 best medium for qualitative purposes, the asparaginate agar for 

 quantitative work. 



CONCLUSIONS. 



Determinations of the number of bacteria present in soil are 

 generally made by counting the colonies that develop on gelatin 

 or agar media. Results depend largely upon the composition 

 of the medium. Three important characteristics are to be looked 



