July 24, 1891.] 



SCIENCE. 



51 



A second possibility of misinterpretation lies in the reduc- 

 tion of the nitrates that may be present in the solution. This 

 reduction takes place even when the quantity of nitrate and 

 organic nitrogen is small, although more slowly than is the 

 case in the presence of considerable quantities of organic ni- 

 trogen. In one example there were no nitrites and .036 

 nitrates present at the beginning of the experiment in the 

 sterilized solution. On inoculation with a certain bacterial 

 species, afterward found to possess a reducing action, the 

 quantity of nitrogen as nitrite increased in a short time to 

 .0256, while the nitrate diminished to .015. On another 

 occasion, with .036 initial nitrate, the nitrites rose from 

 nothing to .021, and the nitrates disappeared proportionally. 

 If larger amounts of nitrate are present, the increase of ni- 

 trite is more striking. Certainly this reducing action of 

 many species of bacteria will go far to explain such results 

 as those reached by Herseus {loc. cit.). 



An interesting experience, and one very significant in the 

 light of our further investigations, should here be mentioned. 

 A nitrogenous solution prepared in the usual way was inoc- 

 ulated with a certain species, — Bacillus ubiquitus, — and 

 examined from time to time, both chemically and bacteri- 

 ally. The solution, on standing for several months, nitrified 

 completely, and the gelatine plate culture showed the pres- 

 ence of a pure culture of B. ubiquitus. We naturally con- 

 cluded that we had discovered a nitrifying organism; but 

 repeated inoculations with a culture of this same organism, 

 both from the flask that had nitrified and from the ori^nal 

 growth in a test-tube, gave a negative result. No better 

 success was had with the same organism freshly isolated 

 from water or soil. No explanation of this perplexing oc- 

 currence could be given at the time, but subsequent events 

 made it probable that our assumed pure culture was not a 

 pure culture at all, but a mixture of the nitrifying organism 

 and B. ubiquitus. Whether the nitrifying organism was 

 introduced from the air, or, as seems more likely, accompa- 

 nied the first inoculation with B. ubiquitus, is unknown. 

 Possibly some of the investigators who have claimed a posi- 

 tive result with species of bacteria grown on gelatine may 

 have been misled in a similar way. 



There was, as has been intimated, one possible explanation 

 of our failure to reach consistent positive results by the use 

 of species of bacteria isolated by the method of gelatine plate 

 culture. It might be that the nitrifying organism did not 

 grow on gelatine. Everything seemed to point in this direc- 

 tion, and the belief was further strengthened by a very sig- 

 nificant fact observed about this time. We had known for 

 some time that ia the history of the filter tanks at the Law- 

 rence experiment station speedy nitrification was always co- 

 incident with a marked decline in the numbfers of bacteria. 

 The effluents discharged from the filter tanks, although high 

 in nitrates, were low in bacteria; and, moreover, the more 

 complete the nitrification, the fewer were the bacteria in the 

 effluent. 



We also observed, that, in an ammouiacal solution which 

 is seeded with ordinary pond water containing several spe- 

 cies of bacteria, there is during the first few days a rapid 

 multiplication of the contained germs. Nitrification, how- 

 ever, does not as a rule begin until from ten to fourteen 

 days have elapsed. By the time nitrification begins, the 

 numbers of bacteria, as shown by gelatine plate cultures, 

 have begun to decline; and, while the nitrogen in the form 

 of nitrites in the solution is increasing, the numbers of bacte- 

 ria are steadily diminishing. Thus, in one instance, an am- 

 nioniacal solution, four davs after its inoculation with a cubic 



centimetre of Cochituate water, contained 3,762,000 bacteriaper 

 cubic centimetre. Nitrification had not yet begun. When 

 the first signs of increasing nitrites appeared, the numbers of 

 bacteria had sunk to 19,200; and when the nitrites reached 

 their maximum, the bacteria, shown by gelatine plate cul- 

 tures, were only 9,454. It was certainly difficult to under- 

 stand why nitrification, a process apparently dependent upon 

 the life and activity of bacteria, should seem to flourish best 

 under conditions in which bacteria were perishing. If, 

 however, it were assumed that the nitrifying organism could 

 not grow in the usual gelatine media, all the perplexing re- 

 sults above recorded could be more easily explained. Under 

 these circumstances it was natural for us to make such an 

 assumption. 



There was, of course, the possibility that the nitrifying or- 

 ganism, by its growth on gelatine, had lost its peculiar 

 property; but it did not seem to us likely that so fundamen- 

 tal a property could be parted with in so short a time. How- 

 ever that might be, we determined to test the other hypothe- 

 sis first, since we believed it to be the more probable of the 

 two. Accordingly, experiments were begun to attempt to 

 isolate the nitrifying organism by the method of dilution. 

 This is the method that was commonly used by investigators 

 in bacteriology before the invention of solid culture media. 

 It has, as is well known, serious practical as well as theo- 

 retical drawbacks. In our practice a small portion of an 

 actively nitrifying solution is transferred on the loop of a 

 sterilized platinum needle to a sterilized ammoniacal solu- 

 tion, and when nitrification is thus induced in the second 

 solution a fresh transfer is made from this to a third, and so 

 on. Rigid precautions have been taken to avoid the intro- 

 duction of foreign germs. 



Hardly were these experiments well under way, before 

 our interest in this method of procedure was stimulated by 

 the publication of communications by Percy F. Frankland 

 and Grace Frankland, and by Robert Warington (Chemical 

 News, Vol. LXI., p. 135). 



The Franklands, having reached a conclusion similar to 

 our own regarding the behavior of the nitrifying organism 

 in gelatine, had also attempted to isolate the nitrifying or- 

 ganism by the dilution method, and had succeeded in this 

 attempt. They state, in their abstract of the paper read be- 

 fore the Royal Society, that, " after a very large number of 

 experiments had been made in this direction, the authors at 

 length succeeded in obtaining an attenuation consisting of 

 about 1 oTTWiTo of the original nitrifying solution employed, 

 which not only nitrified, but, on inoculation into gelatine- 

 peptone, refused to grow, and was seen under the micro- 

 scope to consist of numerous characteristic bacilli, hardly 

 longer than broad, which maybe described as bacillo cocci." 



Warington's communication entirely confirms that of the 

 Franklands, in so far as it relates to their earlier and nega- 

 tive results. He had not, however, at the time of writing, 

 succeeded in isolating the nitrifying organism. 



A paper by Winogradsky followed soon after. He appears 

 to have discovered independently a nitrifying organism, and 

 attributes his success largely to his microscopic examinations 

 of the nitrifying solutions, and to his use of solutions devoid 

 of organic matter. The following is the composition of the 

 liquid adopted by him: ammonium sulphate, 1 gram; potas- 

 sium phosphate, 1 gram ; water from the lake (at Zurich) 

 1,000 grf.ms. Each portion of 100 cubic centimetres received 

 in addition .5 to 1 gram of basic magnesium carbonate, sus- 

 pended in distilled water. Winogradsky found that this 

 layer of magnesium carbonate at the bottom of each flask 



