COMPOSITION OF NATURAL FATS AND OILS 227 



In further support of the theory that a liiohydrogenation may take place 

 m vivo, Hilditch and Stainsby-^^ have shown that, on progressive hydro- 

 genation of lard originally having a low stearic acid content, the proportions 

 of fully saturated triglycerides correspond to those of fats originally having 

 a correspondingly higher stearic acid content. The proportion of GS3 to 

 total saturated acids during the course of hydrogenation of the lard is in- 

 dicated by the solid line in Figure 3. 



In his criticism of the random theory, Hilditch^^ points out that even 

 though it may appear to apply in some cases to the saturated acids, there 

 is no evidence to indicate its applicability to the unsaturated acids either in 

 the animal or in the vegetable kingdom. The highest amount of triolein, 

 or completely unsaturated triglyceride (GU3) , reported in the above fats is 

 3%, whereas the proportion calculated on the basis of U^ would be 2.7, 6.4, 

 12.5, 21.6 and 34.3%, respectively, when the unsaturated acids comprise 

 30, 40, 50, 60, or 70% of the total (or a saturated acid content of 70, 60, 

 50, 40 and 30%, respectively). In a recent study of glyceride structure 

 based on ox, sheep, and pig body fats, Hilditch^^* states that no correlation 

 exists between the proportions of the di- and the triunsaturated glycerides 

 with a random distribution, although the ratios of the trisaturated and 

 monounsaturated triglycerides could be fitted into such a scheme. Hil- 

 ditch-^* sees no reason to postulate a different enzyme synthesis for animal 

 and for vegetable fats. 



There are a number of cases in which Hilditch agrees that the proportions 

 of the four types of glycerides (GS3, GS2U, GSU2, GU3) agi'ee with those 

 calculated on a probability basis. This approximation to random distri- 

 bution, however, may actually be accidental. For example, when the 

 proportions of saturated or of unsaturated acids are relatively high, the 

 values calculated for the trisaturated glycerides, either on the basis of 

 "even" distribution or on the random theory, become almost identical. 

 Experimental errors in the determinations are sufficient to render it im- 

 possible to state which of the two patterns is operative. This is the case 

 with the triunsaturated glycerides of pig back fat^" and Indian sheep fats,^'* 

 as well as Indian cow fats. ^^* 



It is likewise impossible to draw conclusions from experimental values as 

 to the pattern of assembling the triglycerides of seed fats in which the satu- 

 rated acids comprise 80 to 85% of the total acids. Jackson and Longe- 

 necker-^^ have suggested that babassu fat, which is a seed fat belonging to 

 the Pahnae family, follows the random pattern. Hilditch,^^ on the other 

 hand, points out that with a fat such as babassu, which contains 86.7% of 

 saturated acids, one may as well describe the fatty acid arrangement as due 



2" T. P. Hilditch, /. Am. Oil Chemists' Soc, 26, 41-45 (1949). 



M6 T. P. Hilditch and R. K. Shrivastava, J. Am. Oil Chemists' Soc, 26, 1-4 (1949). 



»8 F. L. Jackson and H. E. Longenecker, Oil & Soap, 21, 73-75 (1944). 



