1919] AGRICULTUEAL, CHEMISTRY AGROTECHNY. 207 



flask H, to prevent any acid from getting into the scrubber I, filled with 30- 

 mesh zinc amalgamated with mercury. The trap H may be partially filled 

 with a concentrated solution of silver sulphate to remove chlorine which might 

 escape absorption in I. The upper half of J is filled with calcium chlorid 

 and the lower half with phosphoric anhydrid, the two being separated by a 

 layer of glass wool which is also used at the lower openings of all three 

 scrubbers. The air current after leaving J passes through the U tube K filled 

 with calcium chlorid and into the Nesbitt carbon combustion bulb L, filled 

 with 20- to 30-mesh soda lime except for a small space at the top iu which is a 

 thin layer of phosphoric anhydrid. The U tube M, filled with calcium chlorid 

 and soda liiue, prevents contamination of the combustion bulb by stoppage of 

 the pump. The whole system is connected mth a suction pump. 



From 5 to 15 gm. of soil and 5 cc. of the chromic acid solution are intro- 

 duced into the combustion flask. Fifty cc. of sulphuric acid is run into the 

 flask rapidly and sufficient heat applied under the flask to bring the mixture 

 to the boiling point in five minutes. The heating is continued for 20 minutes, 

 during which the aspiration should be so regulated that about 5 liters of air 

 are drawn through the apparatus in the entire time. The aspiration is con- 

 tinued for five minutes after the heat is removed. When cool the absorption 

 bulb L is closed and weighed. 



Tables are given sliowing the accuracy of the method as determined by du- 

 plicate analyses and by comparison with other methods. The average error in 

 weighing the combustion tube was about 0.02 per cent in terms of carbon, and 

 the determination itself was subject to about the same error. The total time 

 for the determination is about 25 minutes. 



Analysis of calcium cyanamid with large percentage of dicyandiamid or 

 urea, I. E. Hene and A. van Haaeen (Ztscht: Angew. Chem., 31 (1918), No. 53, 

 Aufsatztcil, pp. 129-131). — The method described is said to avoid the error of 

 Caro's method (E. S. R., 25, p. 24) in abnormal samples of calcium cyanamid 

 containing large proportions of dicyandiamid and iirea. The technique of the 

 proposed method is as follows : 



The sample, the nitrogen content of which should be about 0.2 gm., is dis- 

 solved in water, diluted to about 200 cc, and precipitated with 20 cc. of 10 per 

 cent silver nitrate and 30 cc. of 10 per cent potassium hydroxid. The precipi- 

 tate, which contains all the cyanamid and dicyandiamid and some of the urea, 

 is washed, the filter transferred to 150 cc. of water, and the silver salt dissolved 

 iu the smallest possible amount of nitric acid and diluted to 200 cc. Half of this 

 solution is precipitated with 2 cc. of 10 per cent silver nitrate with the addition 

 of 15 cc. of 10 per cent potassium hydroxid. The nitrogen in the precipitate, rep- 

 resenting the cyanamid and dicyandiamid, is determined by the Kjeldahl proc- 

 ess. The urea nitrogen is estimated in the combined filtrates or, more simply, 

 by difference after determining the total nitrogen. In the other half of the 

 original solution of the silver salt the cyanamid is precipitated by silver nitrate 

 and ammonia. The dicyandiamid is calculated by difference. 



The red iodotannic test, D. E. Tsakalotos and D, Dalmas {Bui. Soc. Chiin. 

 France, 4. ser., 25 {1919), No. 2, pp. 80^8^). — ^A further study of the iodotannic 

 test previously noted (E. S. R., 40, p. 610) is reported. With an N/100 solution 

 of iodin the test was found to be more delicate with the tannic reagent than 

 with starch excei)t in the presence of potassium iodid, which in amounts larger 

 than 1 per cent completely inhibits the tannic reaction. 



The amounts of iodin and tannin were found to have a marked effect upon 

 the reaction. The maximum intensity is produced when the quantities of iodin 

 and of tannin are in the proportion of 1 cc. of N/10 iodin to from 0.4 to 1.2 of a 

 1 per cent tannin solution. 



