2 BULLETIN 103, U. S. DEPARTMENT OF AGRICULTURE. 



each case tests were made on healthy young men by including 

 aluminum in some form in their food. The food was all carefully 

 measured and weighed and the amounts of its principal ingredients 

 were determined by analysis. The excretions of the men's bodies 

 (both urine and feces) were carefully collected, examined, and ana- 

 lyzed. Daily records of body weight, temperature, respiration, and 

 pulse were kept for each man, and notes were made of any unusual 

 symptoms. Any disturbance in health or physiological processes 

 was thus detected. 



Each experiment included three periods, in the first and last of 

 which no aluminum was administered. During the middle period 

 aluminum compounds were administered, the "dose" increasing as 

 the experiment progressed. In this way the effect of large quanti- 

 ties was compared with that of small quantities. In Dr. Chitten- 

 den's and Dr. Taylor's experiments some of the men who served as 

 " control" subjects received no aluminum at any time, so that any 

 disturbances due to other causes might be checked up. 



Dr. Chittenden's experiments included 12 men and continued 

 from January 15 to June 22, 1912. During 130 days the diet con- 

 tained bread raised with an alum baking powder made in the labo- 

 ratory. 1 The dose of aluminum compound was increased from time 

 to time, at first by increasing the quantity of bread and later by in- 

 creasing the quantity of the baking powder used in making the bread. 

 In this way the alum 2 used per man per day was increased from 

 0.578 gram 3 (8.920 grains) at the beginning to 2.287 grams 4 (35.295 

 grains) at the close of the dosage period; the actual aluminum con- 

 tained in this dosage ranged from 0.065 gram (1.003 grains) to 0.257 

 gram (3.966 grains) per man per day. Eight men used the alum 

 bread, while four had no aluminum in their food. 



Dr. Long's experiments ran from February 8 to June 7, 1911, and 

 included six men, all of whom received the dosage. Baking powder 

 bread was not used, but instead for 40 days a mixture of the same 

 composition as the residue left in such bread by alum baking powder 

 was administered in the form of a powder in water or milk. For 30 



1 This bread was made fresh every day and contained in one baking of two loaves approximately: 



Sift ed flour quarts . . 2 



Baking powder (25 per cent calcined alum) heaping teaspoonfuls. . 4 



Salt (approximately one rounded teaspoonful) ounce. . J 



Butter do 1 



Water, sufficient quantity. 

 Laterin the experiment a greater proportion of alum baking powder was used in the making of the bread 

 in order to facilitate administering larger amounts of alum. 



2 The term "alum" as used under the heading " Character of experiments conducted ' refers to the cal- 

 cined sodic aluminic sulphate commonly used in alum baking powders and not to the ordinary crystallized 

 alum. 



3 Equivalent to approximately two-thirds of a level teaspoonful of baking powder containing 25 per cent 

 ofalum. All the figures in this and succeeding footnotes must of necessity be approximate, since teaspoons 

 vary in size and baking powders in composition. 



* Approximately equivalent to 2§ level teaspoonfuls ofalum baking powder. 



