122 BACTERIA IN RELATION TO PLANT DISEASES. 



the inoculating materials and the methods of inoculation, the cases where results of inoculation are 

 completely lacking will be exceptions." 



The third group of 25 experiments all showed positive results from the inoculation. One of these 

 reported by Gerlach was peculiar in that while the green substance showed an increase in weight, 

 the weight of the dry substance fell below that of the uninoculated plants. These plants were subject 

 to a long drought during the middle period of their growth. Hiltner observed the same phenomenon 

 with soy-beans which ripened late, though under favorable weather conditions the growth of the soy- 

 bean is increased by the presence of nodules. 



In one set of experiments it was noticeable that the small, seeded species gave good results from 

 seed inoculation while with large seeded species such as peas and lupins, the results were either 

 doubtful or negative. This was probably due in the case of the large seeds to the inhibiting action 

 of excretions from the seed-coat. 



Another grower obtained a gain of 50 per cent in straw and seed with soy-bean. Hiltner who 

 saw and examined the mature crop early in October, states that the roots of the uninoculated plants 

 were free from nodules, while nodules were formed on about 50 per cent of the plants in the inoculated 

 plots. 



In another case where poor soil was used the inoculated vetches were backward compared to 

 the uninoculated ; later, however, they outstripped them with a more luxuriant green and a denser 

 mass but the weeds also seemed to grow better there. 



Director H. Rose obtained striking results with serradella and yellow lupin. The seeds were 

 carefully inoculated and sowed in good weather. The field used was a new, light sandy, rather moist 

 heather land, which had borne previously only rye, fertilized with lime-kainit-phosphate and Chile 

 saltpeter. 



For this experiment 500 kg. of Thomas-meal and 600 kg. of kainit per hektar was applied in the 

 spring. The seeds were sown on May 1 1 . The weather was very dry, only two rains occurring during 

 the whole summer. Growth which began slowly in the serradella took on new life late in June in the 

 inoculated plots, so that early in July while the plants of the uninoculated plots were barely 10 cm. 

 high, unbranched, yellow, and sickly with absolutely no nodules, those in the inoculated plots were 

 20 to 30 cm. high, branched so as to form a close mat and with roots full of thick, watery nodules. 

 The results with lupin were similar but less marked. The harvest of serradella (Ornithopus saliva) 

 is given below : 



Plot. 



Green weight. 



Seed. 



I. Uninoculated, per hektar. .. 5,300 kg. 410 kg. 



III. Uninoculated, per hektar. ..! 6,800 kg. 470 kg. 

 II. Inoculated, per hektar : 12,250 kg. 680 kg. 



IV. Inoculated, per hektar i 11,750 kg. 615 kg. 



Hiltner also saw the crops of Tacke, Director of the Moor Station at Bremen, just before they 

 were harvested, and states that of the uninoculated plants, most were suffering hunger. There were 

 yellow lupins, blue lupins, scarlet clover, and serradella. A few plants showed by their growth the 

 presence of nodules, but others (yellow lupin and scarlet clover) possessed large nodules which were 

 completely inactive. On yellow lupin, seed-inoculations gave no improvement over the check but 

 inoculation with infected soil gave a luxuriant crop, 80 cm. high, all plants having nodules which were 

 arranged in rows mostly on the lateral roots. 



Blue lupin gave similar results. About 30 per cent of the uninoculated plants formed on the 

 deeper roots a few nodules which exerted little influence on the growth. Seed-inoculation gave 35 

 per cent of nodule-bearing plants which, however, did not show any benefit from the nodules, which 

 were on the lateral roots and hence due to spontaneous infection. Inoculation with infected soil 

 produced luxuriant growth, all the plants bearing numerous nodules. Inoculation with natural earth 

 gave equally good results. 



Uninoculated crimson clover was extremely scanty and sickly in growth in spite of nodules on 

 the roots of all the plants. Here seed-inoculation gave the best results, with numerous nodules on 

 all parts of the roots. 



Uninoculated serradella was also sickly. Seeds which were soaked before sowing gave by far 

 the best results. 



These experiments brought out the fact that in all cases uninoculated plots, or plots inoculated 

 with pure cultures, were almost entirely free from weeds, while those on which natural soil inocula- 

 tions were made were overgrown with weeds, especially wild spurry which is hard to eliminate. This 

 observation has been repeatedly made by Tacke. 



