342 Journal of Agricultural Research voi. xviii, no. 6 



On the other hand, there evidently has been a gain in weight of composts 

 No. 3 and 4 on account of sulphofication, and we would expect some loss 

 in nitrogen from the manure during decomposition. 



Unfortunately these samples had previously been kept in the labor- 

 atory in Mason jars with the tops on loosely so that air was not exlcuded, 

 and they may have taken up ammonia fumes. But if such is the expla- 

 nation it would seem that the acid composts No. 3 and 4 would have 

 absorbed more than the others. The respective gains in No. i, 2, 3, and 

 4 were 0.0536, 0.0604, 0.0659, ^.nd 0.0349 per cent. 



On the other hand, the composts from which the air was excluded as 

 well as those exposed to the weather showed losses of nitrogen; but these 

 losses may have been caused by denitrification in the former and leaching 

 in the latter. The fact that leaching apparently did not cause a loss of 

 nitrogen in No. 3 and 4 greater than that caused by the exclusion of air 

 indicates that denitrification took place in the samples from which the air 

 was excluded. 



In the later experiments, after 17 months' time, about 27 per cent of 

 the phosphorus present had been made citrate-soluble, as shown in Table 

 VIII. In the earlier experiments, in about the same time, 72 per cent 

 was citrate-soluble. After about 8 months, however, nearly 1 1 per cent 

 of the total phosphorus was citrate-soluble in these experiments as com- 

 pared with 4 per cent in the earlier ones. This was probably due to 

 better conditions for controlling temperature, although part of the initial 

 good effect was undoubtedly due to the starter. 



The results show that the reaction proceeds more rapidly at a high 

 temperature and when the composts are stirred frequently. While No. 

 6 and 10 at the end showed about the same amounts of citrate-soluble 

 phosphorus, the effect of temperature is illustrated during the first 

 months of the experiments. In fact, it required the heat of the summer 

 months before pronounced activity was shown in any experiment. 



Because much less soil and no manure was used in the latter work, 

 the nitrate figures are not so significant as in the other experiments. 

 Nevertheless, as much nitrate was present at higher acid concentration 

 as was found at the begirming, or more. 



That the presence of the sulphof ying organism in the amount of starter 

 used is not sufficient to continue the reaction as rapidly as does the 

 further addition of soil bacteria can be seen in Table XII. Another fact 

 to be considered in this connection is that increasing the amounts of soil 

 sHghtly favored the reaction, although soil water obtained from a large 

 mass of the same soil gave about as good results. 



The sulphofying power of different types of soil varied somewhat, but 

 all had the capacity to oxidize sulphur. The relative amounts of sulphur 

 oxidized in individual cases did not vary materially, although the totals, 

 while small, differed as much as 100 per cent. None of these soils, 



