358 Journal of Agricultural Research voi. xiii, No. 7 



The color and characteristic odor of the oil seemed to diminish gradually 

 in the control samples. The heated samples, series B, duplicated the 

 control samples, so far as could be observed. 



Air was a negligible factor for two years; then it effected a slow but 

 marked destruction of color fully equal to light at the close of the experi- 

 ment, and caused a rancid odor on the sixth year. 



Light was active in destroying color and caused a slightly rancid odor 

 on the sixth year, probably due to a small amount of inclosed air. 



Moisture caused the formation of a precipitate which rendered the oil 

 turbid, but which effected no apparent change in color after the removal 

 of the precipitate. 



Air-light was most active and effective in destroying color, equal to 

 air-light-moisture, and produced a rancid odor on the second year. 



Air-moisture caused the formation of a slight amount of precipitate 

 but without appreciable turbidity until the fifth year, at which time a 

 rancid odor was produced. Air-moisture was inactive as regards color 

 for three years, but eventually exceeded the effect of air and equaled 

 that of light. 



Light-moisture affected the color about the same as light, and caused 

 the formation of a considerable amount of precipitate which rendered 

 the oil turbid. 



Air-light-moisture affected the color the same as air-light, produced a 

 rancid odor the second year, and caused the formation of probably the 

 largest amount of precipitate, which rendered the oil turbid. 



The chromogenic bodies of the oil were not appreciably affected by 

 moisture, were destroyed slowly but effectively by air, slowly but rather 

 more effectively by air-moisture, more actively by light and light- 

 moisture, and most actively and effectively by air-light and by air-light- 

 moisture. Air was slowly active, light probably assisted by a small 

 amount of inclosed air more active, and air-light the most active in de- 

 stroying color. Moisture was a negligible factor except possibly in the 

 case of air-moisture. 



A rancid odor was produced on the sixth year by air and by light, on 

 the fifth year by air-moisture, and on the second year by air-light and by 

 air-Hght-moisture. Neither air nor light alone was particularly active 

 in producing rancidity, but jointly were decidedly effective. In this 

 connection moisture did not appear to be a factor of any consequence. 



In every instance the presence of moisture caused the formation of a 

 precipitate in a relatively .slight amount by air- moisture, in small amount 

 by moisture, in a greater amount by light-moisture, and in apparently 

 the largest amount by air-light-moisture. Light seemingly was a factor. 

 The so-called precipitate was first observed as dirty-white or brownish- 

 white spots on the sides of the bottle below the surface of the oil and 

 might be said to resemble mold. As the amount increased, the bulk of it 

 collected near the surface of the water layer or in the water. In no case 

 was sufficient purified material obtained to make a chemical examination. 



