Oct. 8,1921 Relation of Sulphates to Plant Growth and Composition 103 



volatile matter was determined by ashing a ground sample representing 

 each cultm-e. This was done to correct for any excess weight due to 

 adhering soil particles. 



The total sulphur was determined by the sodium-peroxid method. 

 The sulphate sulphur was extracted by taking 2 gm. clover and 150 cc. 

 of water and digesting on the steam bath for three hours. It was then 

 slightly acidified with hydrochloric acid, and after standing for an hour 

 the extract was filtered. Five cc. of 10 per cent barium-chlorid solution 

 was used to precipitate the sulphate sulphur in the hot solution. After 

 standing overnight the barium sulphate settled to the bottom of the 

 beaker in all cases, and no particles of precipitate could be detected in 

 the supernatant liquid. This liquid was carefully decanted off; and the 

 white precipitate was washed on a Gooch crucible, dried, and weighed. 

 Several of these precipitates were ignited, but since no appreciable loss 

 in weight was detected this method appeared perfectly reliable for 

 comparable results on sulphate sulphur in the different samples of plant 

 material. For the determination of total nitrogen insoluble in acetic 

 acid, i-gm. samples were digested with about 150 cc. of water on the 

 water bath for two hours. The extract was then acidified with dilute 

 acetic acid and filtered after standing about 30 minutes. The total 

 nitrogen was determined on the precipitate by the Kjeldahl method. 

 The filtrate was made alkaline with sodium carbonate, evaporated to a 

 few cubic centimeters in volume, transferred to a nickel crucible, and 

 total sulphur determined by the sodium-peroxid method. 



The dry weights of plant material produced show no increase in pro- 

 duction that can be attributed to presence of sulphates. This result is 

 different compared to the noticeable increase reported with the same 

 soil before (4). As mentioned above, conditions were very unsatisfactory 

 for growth, and the plants were cut before maturity five months after 

 planting. In the former work reported, conditions permitted rapid 

 growth, and the plants, though not mature, were harvested two months 

 after seeding. As the soil used in this later work had been heated, there 

 was perhaps some change in degree of solubility of soil minerals and in 

 the biological flora. The concentration of added mineral salts was also 

 greater in this work. 



Examination of the roots showed that all plants had become infected. 

 In the uninoculated series, roots from cultures i and 2 contained very 

 few nodules compared to the roots grown in the soil receiving sulphate 

 fertilizer. This remarkable difference in nodule formation no doubt 

 accounts for the low nitrogen content of the clover plants in pots i 

 and 2. That these plants became infected without any artificial inocu- 

 lation is not surprising. Wilson (8) found that — 



of fifteen legumes grown in Volusia silt loam soil, only one, Trifolium pratense, 

 developed nodules without artificial inoculation. 



