132 EXPERIMENT STATION RECORD, 



In addition to these experiments, 3 small plats were planted with alsike 

 cIoA-er and 3 with field peas, and inoculated by mixing the cultures with steril- 

 ized skimmed milk and moistening the seeds with this mixture. The results 

 showed conclusively that inoculation did not benefit either the clover or the 

 peas, but rather the reverse. In all the plats tubercles were present on the 

 roots and were as abundant on the uninoculated as on the inoculated. 



A third series of experiments was conducted with peas and clover on land 

 which had not grown legumes for a number of years (in some cases 10 years), 

 in which half of the plats were dressed with lime at the rate of 2 tons per acre. 

 Both crops were harvested when the plants were in full flower. A general 

 average of the yields from inoculated and uninoculated plats showed no mate- 

 rial results from either inoculation or liming. Selected plants lifted from each 

 plat and compared as to the root system revealed tubercles on all of them, 

 being very abundant and large on the peas, while those on the clover were 

 small, few in number, and brown and shrivelling at the time the examination 

 was made. The author claims as the net results of these experiments that 

 root tubercle bacteria are able to exist in soil for at least 10 years. 



In soils containing these bacteria, even if they are not abundant, no appre- 

 ciable benefit will be derived by inoculating the seed or soil, as the number 

 so added will be trivial as compared with those already present. If after a 

 careful examination of the plants grown on any soil it is certain that tubercle 

 bacteria are lacking, the surest method to obtain them is to inoculate the new . 

 soil with 130-300 lbs. per acre of soil from an old field that previously had 

 grown legumes. If this is not available, the root tubercles may be stripped 

 from plants, pounded to a thin paste with water, and mixed with the seed 

 before planting. Xitragin or other nitro-cultures may be used, provided the 

 cultures are fresh and suitable for the particular crop, but this is the most 

 expensive and the least efl'ective method of infecting sterile soil. 



New infection experiments with nitrobacterine, nitragin, and soil cul- 

 tures on blue lupine, H. von Feilitzen (Centbl. Bakt. [etc.], 2. Abt., 26 {1910), 

 No. 10-12, pp. 3.'i5-3o2, figs. .'/). — The results are given of infection experiments 

 with nitrobacterine, nitragin, and soil cultures on blue lupine grown on high 

 moorland under cultivation. 



It was found that legumes grown on this type of soil for the first time did 

 not produce satisfactory results, unless they were infected with cultures of 

 tubercle bacteria. Soil cultures from fields which had iireviously grown legumes 

 produced the surest I'esults and the highest yields. So long as the land had 

 grown a crop of legumes the previous year, the source of the soil cultures was 

 immaterial, soil cultures from vetch and clover land producing satisfactory results 

 on seradella and lupine. The results with nitragin were not so satisfactory as 

 with soil infection, while the nitrobacterine under the conditions was entirely 

 without value. 



Mutual interaction of plant roots, J. B. Dandeno (Rpt. Mich. Acad. Sci., 

 11 (1909), pp. 2.'/, 2-'}, pJ. 1). — In continuation of previous experiments (E. S. R., 

 21, p. 319) the author has conducted additional investigations on the mutual 

 interaction of plant roots. 



It has been noticed that Canada thistles in grain fields were found growing 

 in places where the best grain occurred, and this led to a series of pot experi- 

 ments in which the effect of the underground system of the Canada thistle on 

 the growth of oats, barley, buckwheat, wheat, and flax was studied. With the 

 oats the thistles seemed to do no harm, but rather tended to increase the crop. 

 The buckwheat showed exactly the opposite result. Tweuty-two days after 



