It will also be seen t h a t there is a variation in the results of the two different men using the same soils and Rather method, being a difference of 0.37 per cent. in soil 12-1-110. I n only one soil, 12-1-143, do the results agree. This seems to indicate that if the exact amount of ammonium carbonate is not added each time and the solutions do not have exactly the same treatment, different results are obtained. The per cent. of humus obtained by Chemist No. 3 was slightly higher than that

more »

... tained by Chemist h-0. 2 ; also the per cent. of ash was higher. But this is easily explained by the fact that this determination was the first time Chemist No. 3 had used this method. The solutions from which he drew his aliquot gave a slight deposit of clay on standing. He did not allow the filtrate to become perfectly clear before he took the portion for the determination. I n our routine analyses, we allowed about fifty of these solutions t o stand two weeks and at the end of t h a t time only two or three showed any precipitate at all. The greatest difference between the results of the two men doing the proposed method is 0 . 1 4 per cent. in soil 12-1-122, notwithstanding the fact that one of the men had never used the method before. S U M M A R Y I . The weighing out of an exact amount of am-2. While conducting the analytical work on the ash of cow feces, in a mineral nutrition experiment, it was found that the results were not concordant. This is due to the fact that the ash of COTV eces runs abnormally high. The ash content of the feces used in this investigation was 2 0 . j j per cent.: while Dr. P. F. Trowbridge,Z of the Missouri Experiment Station, has found the ash to run as high as 39 per cent. This unusually high ash content is due to the large amount of sand, dirt, etc., that the cow receives nTith the feed. Methods of ash analysis are very lax with reference t o the acid treatment of the ash. The time varies from I O minutes3 boiling with acid, for plant ash; t o I O hours4 digestion over steam, for soils. As nothing could be found relating to the ash of cow feces, i t was decided to determine the factors which influenced the removal of the mineral constituents. ' Abstract of thesis suhmitted in partial fulfilment for Degree of Master of Arts, Cniversity of Xlissouri, 1912. Private communication. "Food Inspection and Analysis," Leach, p. 303. Bullefin 107, Official and Provisional Methods, U. S. Dept. of Agriculture. The preliminary work showed that the calcium was the constituent removed with the greatest difficulty and therefore might serve as an index to all of the other mineral constituents. The factors which i t seemed necessary to determine may be outlined as fol1on-s:-I . The time necessary to remove the lime by. (a) digestion of ash with acid over steam, ( b ) boiling the ash with acid on a hot plate. 2 . The comparison of the efficiency of dilute and concentrated acid. ( I t nTas finally concluded t o use concentrated hydrochloric acid, concentrated nitric acid and dilute hydrochloric ac d of specific gravity r.11.1 This dilution is made by mixing equal parts of concentrated hydrochloric acid and water.) 3. The determination of which acid (nitric or hydrochloric) is most efficient. 4. The effect of dilute sodium hydroxide to break up silicates not decomposed by acid. The efficiency of moist combustion as compared with ordinary combustion. The feces were ground to a fine powder when dry and preserved in air-tight bottles. A large portion of the powdered sample was ashed t o a cream white ash and preserved for analysis. The moist combustions were conducted in Kjeldahl flasks, the boiling of the ash was conducted in Erlenmeyer flasks and the digestions were made in beakers, all of which were made of standard glass. Watch glass covers (on the beakers) and test tubes of water (suspended in the flasks) acted as condensers, and prevented loss of acid by evaporation. After treatment with acid, the sample was diluted to twice its volume t o prevent the acid from attacking the filter paper, and the sample was filtered and washed. The filtrate was then evaporated to dryness. The silica residue was taken up with dilute acid, filtered and n-ashed. This filtrate was made up to volume for analysis. Sliquots were made alkaline with ammonium hydroxide and brought back to faint acidity with acetic acid. After boiling a few seconds, the insoluble phosphates were filtered off and washed thoroughly. The calcium was then precipitated as the oxalate and determined with N/ j o potassium permanganate in the usual manner. The following data shoxs (I) that the lime is removed in a shorter period of time when the ash is boiled with acid than \?-hen it is digested over stcam. ( 2 ) That the ash must be boiled three hours to insure complete removal of the lime. (3) That in digestion, concentrated acid is more efficient than t h e acid diluted I : I . (4) That when boiled, the concentrated acid has no advantages over the weaker acid. When the acid is incapable of breaking up the silicates, it is advisable to treat the large silica residue with sodium hydroxide as proposed by the Fresenius-Will method.' When these alkaline washings are * This is the strength of acid used in soil analysis (BzLU 107, U. S. Dept. 2 Jour. I. Gasbeleuchtung, 49, 853 (1906). of Agriculture).