CHEMICAL AND BACTERIOLOGICAL EXAMINATION OF SOIL. 761 



are placed in tin incubating cases, in which they rest on filter-paper, kept moist with 

 corrosive sublimate solution. 



These plate cultivations were allowed to incubate as a rule for four days at 16° 

 Centigrade, after which the colonies were counted, examined, and, where necessary, 

 subcultures made from them. On each sample of soil three plates were done. One 

 plate was inoculated with 2 drops of the suspension, another with 4 drops, while the 

 third had 8 drops added to it. In dealing with samples from depths greater than 

 5 feet from the surface, I usually inoculated the plates with 2, 4, and 16 drops 

 respectively. In counting, the mean of the plates was always taken, although in most 

 cases the numbers present in the various plates consisted well with one another. On 

 the later samples a worts-gelatine cultivation was made for the purpose of estimating 

 the number of moulds present. Such cultivations were inoculated with one cubic 

 centimetre of the suspension. 



Details of Chemical Examination. 



Organic matter. — The method used for the estimation of the organic matter present 

 in the samples was Dr Hunter Stewart's modification of the Kjeldahl process, com- 

 municated by Dr Stewart to the Royal Society of Edinburgh in December 1892. Dr 

 Stewart has so modified the process that organic carbon and organic nitrogen can be 

 estimated at the same time, and with great facility. 



la applying the method to soil, it was thought well to first add about 15 cubic 

 centimetres of dilute sulphuric acid to the earth in the flask, aspirate for half an hour 

 before putting the baryta tubes into the circuit, gently heating during that time, allowing 

 the flask to cool, and sweeping out for twenty minutes thereafter. The object of this 

 was to get rid of any carbonic acid which might be present in the form of carbonates. 



The quantity of earth used was, as a rule, 5 grammes for each estimation, although 

 in some of the purer samples 10 grammes was the working quantity. Each sample got 

 exactly the same treatment, and in practice as good results were got when only 2*5 

 grammes were used as when 5 or even 10 grammes were employed. 



In order to prevent bumping and loss of flasks and time, a sandbath, heated by a 

 large Fletcher's burner, had to be used as the source of heat in distilling. 



Free Ammonia. — The ammonia, whether free or in the form of its salts, was esti- 

 mated in the following way. Five grammes of the soil, previously rubbed down in a 

 mortar, were put into a large bolt-headed distilling flask, rendered slightly alkaline with 

 " ammonia" free caustic soda-solution, and distilled. Each 50 c.c. of the distillate was 

 Nesslerised as it came over. The amount of ammonia thus calculated was converted 

 into nitrogen and deducted from the result got by the Kjeldahl method. 



Moisture. — A quantity of the earth, generally about 10 grammes, was rubbed up in 

 a mortar and dried in the hot-air chamber until a constant weight was attained, and the 

 percentage of moisture calculated. 



