table, the non-spore-formers have averaged 75.6 per ct. of all the colonies on the plates. This has been found to be about the general average in cultivated soil; in sod soil (in which, as already stated,11 the number of Actinomycetes is higher than in cultivated soil) the percentage of non-spore-formers is somewhat lower.

The majority of this 75 per ct. is composed of the "slow-growers." The rapid-liquefying group is not a very abundant one in the soils studied. In many cases it is either lacking or not abundant enough to appear on plates in the dilutions used; and the largest number ever found, except in manured soil, is about 1,500,000 per gram of soil or about 5 per ct. of all the colonies developing on the plates. The group of "slow-growers," on the other hand, is very abundant. Excluding air-dried soil, where their numbers are quite low, and manured soil, where their numbers are extremely high, plate counts have ranged from about 2,000,000 per gram to about 60,000,000 per gram.

FLUCTUATIONS IN NUMBERS.

The mere fact that the majority of the bacteria developing on plates made from soil are non-spore-formers suggests that these organisms are important in soil; but still stronger evidence of their importance comes from the fact that they fluctuate greatly in numbers. In this respect they differ greatly from the spore-forming bacteria, which as already mentioned 12 are relatively constant in numbers. The figures in Table I illustrate the fluctuations in their numbers. During the three years over which the sampling was continued the plate count of non-spore-formers has varied from 5,000,000 to 44,000,000 per gram; while all the other organisms growing on the plates (spore-forming bacteria and Actinomycetes) have varied only from 3,200,000 to 10,500,000 per gram.

A further illustration of this same fact can be obtained from the data already published in Cornell Bulletin 338. It will be seen from Table II in that bulletin (p. 87) that the number of colonies of "slow-growers" varied from 2,000,000 to 25,000,000 per gram; while (excluding the first two counts in the table, which as stated in that bulletin are probably inaccurate) the number of all the other colonies varied only from 1,700,000 to 8,800,000 per gram.

Even more striking fluctuations in the number of non-sporeformers have been observed in pot experiments. The aeration brought about by potting soil is generally admitted to cause an

between 50 and 100 colonies per plate (or nearest to this number, in case neither dilution gave plates within these limits). In almost all cases, moreover, parallel plates were made with some agar medium; and altho the counts given in the tables were never obtained from the agar plates, the use of agar sometimes helped to show which of two disagreeing gelatin plate counts was correct.

11 Conn, H. J. A possible function of Actinomycetes in soil. N. Y. Agr. Exp. Sta., Tech. Bul. 52. 1916. Jour. Bact., 1:197-207. 1916.

12 See footnote 3.,

increase in the numbers of bacteria. In the course of the present work it has been noticed that the numbers of spore-formers and Actinomycetes remain about constant after aeration and that only the non-spore-forming bacteria increase noticeably in numbers. Ă good illustration of this is given in Table II. The soil used in this experiment was kept in a pot in the laboratory, watered occasionally so as to keep the moisture content at about 10 to 20 per ct. The soil placed in this pot was obtained from a field plat from which plates GELATIN PLATE COUNTS OF AERATED SOIL.

TABLE II.

Figures indicate number of colonies per gram of dry soil.

had been made at intervals for two or three years, so that the composition of its flora was fairly well known. The first count listed in the table is a typical analysis of this soil before aeration. It will be seen that during the first week or two after aeration the count of rapid-liquefying non-spore-formers rose to about four times its normal height, remaining high during the first month, afterward returning to normal height or even lower. The number of punctiform colonies increased immediately to twice normal, and did not return to its original level until after the first month, and then only temporarily, later on increasing again to more than twice its normal height. Meanwhile the number of colonies of other organisms (Actinomycetes and spore-formers) remained strikingly constant, the lowest count being 2,700,000 and the highest 4,700,000 per gram, and such fluctuations as they did show in their numbers bore no relation to the aeration.

Other experiments have been performed which illustrate the fact that following aeration of the soil or the addition of organic matter, the greatest numerical increases occur in the group of non-sporeforming bacteria. There is no need of giving all the data, but the experiment summarized in Table III is striking enough to deserve mention. In this case two pots were filled with the same soil, fresh horse manure having been added to one but not to the other. The pots were kept in the laboratory and watered as in the experiment just described. In the manured soil the number of rapidly liquefying colonies of non-spore-formers and the number of punctiform colonies each increased to about one hundred times normal during TABLE III.— GELATIN PLATE COUNTS OF AERATED SOIL, MANURED AND UNMANURED. Figures indicate number of colonies per gram of dry soil.

*The count obtained from the original soil is listed under the heading "Manured Soil " as well as under" Unmanured Soil," because this same soil was used in both the manured and unmanured pots.

the first two weeks after manuring, and then gradually subsided, reaching their normal level after six months. Meanwhile the number of colonies of other kinds merely increased to about twice normal; which slight increase might easily have been due to the inaccuracy introduced by the high dilutions necessary. A comparison between manured and unmanured pots, shows plainly that the immense difference between the total counts of these soils was due wholly to differences in numbers of the non-spore-formers.

These data in regard to manured soil are surprising as well as interesting. The spore-forming bacteria and the Actinomycetes have frequently been mentioned as probably taking part in the decomposition of manure; but in this experiment they remained

almost constant in numbers while decomposition of the manure was going on rapidly -as proved by the strong odor of ammonia, if by nothing else. This experiment was originally planned with the idea of furnishing favorable conditions for the growth of sporeformers in soil; and it was with some surprise that they were observed to remain constant while the non-spore-formers increased in numbers enormously. This experiment has not yet been repeated; so the conclusion that spore-formers and Actinomycetes do not take part in the decomposition of manure in soil is not yet justified. The findings suggest that decomposition of manure is brought about by the non-spore-forming bacteria; but they are summarized here merely as an illustration of the fact that they are the organisms which show the greatest fluctuations in number whenever the total numbers of bacteria increase or decrease. 13 Further interesting information as to the occurrence of Ps. fluorescens was obtained from an experiment planned for an entirely different purpose. Some quartz sand was mixed with nutrient material, placed in tumblers and sterilized. After sterilization the tumblers were inoculated in various ways, but in no case with Ps. fluorescens. No attempt was made to prevent contamination with air-borne organisms, as it was desired to see which of the organisms were capable of prevailing over rival organisms under the conditions of the experiment. Fifteen out of 16 tumblers were later found to be contaminated with Ps. fluorescens, which developed to considerable numbers and in some cases persisted for a year. Plainly this organism found conditions in these tumblers well suited to its growth. The exact meaning of this fact is still unexplained; but these results, taken in connection with those given in Tables II and III, suggest that Ps. fluorescens is pre-eminently suited to growth in moist and well aerated soil or sand.

SIGNIFICANCE OF THE NON-SPORE-FORMING
BACTERIA IN SOIL.

It has been emphasized in the past that the most abundant bacteria in soil are not necessarily the most important ones; so the large numbers of non-spore-formers found in soil must not be accepted alone as proof of their importance. The evidence generally considered necessary to corroborate that obtained from numbers alone is to show that they can carry on some activity known to take place in the soil; but evidence of this sort in regard to the asporogenous bacteria has been obtained only in the case of the

13 A further illustration may be obtained from the data to be published in Technical Bulletin 60 (Actinomycetes in soil) Table II. In this case a large increase in the nonspore-forming bacteria was noticed after mixing grass roots with soil.

group of rapid liquefiers. Ps. fluorescens is known to be an ammonifier; and all the members of the group have proteolytic powers. This, together with the fact that they have been found to be especially abundant in freshly manured soil, suggests that they may be the important soil ammonifiers.

The activities of the "slow-growers are more problematic. Some do not liquefy gelatin, while others liquefy it so slowly that the importance of their proteolytic ability is at least questionable. Nevertheless, the increase in their numbers in the soil to which manure was added (see Table III, p. 153) suggests that in this experiment, at least, they took part in some stage of the decomposition of manure. Winogradsky's nitrifiers are non-spore-forming bacteria, and the idea has suggested itself that nitrification might be the function of some of the asporogenous organisms studied in the present work. It is not likely, to be sure, that any of the forms which grow on gelatin plates are nitrifiers, because Winogradsky's nitrifiers did not grow on gelatin; but the possibility is not entirely ruled out of court, for in the present work the gelatin has been of a different composition and the general technic quite different from that used by Winogradsky and others who have studied the nitrifiers. For this reason about ten cultures, isolated from punctiform colonies, were inoculated into a mixture of sand and Omelianski's solution,14 but no signs of nitrite appeared. No definite statement, of course, can be made on the basis of such meager data, and when a more complete study of these organisms is made, nitrifying power is one of the tests that must be included. For the present their function in the soil remains a matter of speculation. It is hoped that this paper will interest others in them, so that they may be given a more thoro study in the future.

A stronger assertion of the importance of asporogenous bacteria in soil could be made if their functions were known; but one fact brought out by the present investigation is fully as good an argument for their importance as a knowledge of their activities in pure culture would be. This fact is that when any change in soil conditions causes an increase in the number of bacteria present, the greatest increase always seems to occur in the group of nonspore-forming bacteria. This plainly indicates that they are active in soil; and considering that they comprise over half of the soil flora, it is hard not to believe that they are important.

The knowledge that they are active in soil is perhaps a better argument for their importance than would be a knowledge of their activities in laboratory culture. Knowledge of the activities of

14 Omelianski, V. Ueber die Isolierung der Nitrifikationsmikroben aus dem Erdboden. Centbl. Bakt., II Abt., 5:537-549. 1899.