The Philippine Journal of Science 



The results of the tests described in the preceding section 

 are recorded in Tables 4, 5, 6, 7, 8, and 9. Instead of discuss- 

 ing the tables separately it will be found more convenient to 

 take up the data of all six tables, column by column. 



Samples.— The lettering or other designation of the samples 

 indicates their correspondence with the samples mentioned pre- 

 viously in this paper. 



Hydnocarpus alcalae, A, was distilled into five instead of six 

 fractions. The fourth and fifth fractions are tabulated with 

 fractions 5 and 6, because they are the last two fractions. 

 Fraction 4 is accordingly lacking for this sample. 



Taraktogenos kurzii, D and F, have small final fractions be- 

 cause of accidental difficulty in washing the crude ester and 

 consequent lack of ester for distillation. 



Pressure.— No appreciable decomposition takes place at 20 

 millimeters. Lower pressures are disadvantageous from the 

 point of view of fractionation. In the cases where the pressures 

 are different from 20 millimeters a rough correction of 1° per 

 millimeter may be applied to the temperatures. 



Temperature. — The temperatures are not corrected for stem 

 exposure. They indicate that between the first and last frac- 

 tions satisfactory separation was obtained, but otherwise are 

 not of much value. Variations due to small leaks and other 

 experimental difficulties mask differences which may be charac- 

 teristic of the individual oils. Hydnocarpus alcalae, B, how- 

 ever, shows the relative constancy of boiling point which agrees 

 well with the later data. 



Weight.— Approximately 50 mils were collected in each frac- 

 tion, but there was some accidental variation, as is shown by 

 the column of weights. The shortages in Taraktogenos kurzii, 

 D and F, Tables 8 and 9, were caused by the inadvertent use 

 of too little ester for distillation. 



Fatty acids; weight.— The weight given in column b for any 

 fraction shows the solubility of the fraction in 200 mils of cold 

 80 per cent alcohol. It was difficult to secure uniform depo- 

 sition in all samples due to such factors as retention of filtrate, 

 supersaturation, re-solution of crystals while filtering, etc. As 

 solubility figures, therefore, these columns are not very accu- 

 rate, but they are necessary in connection with later figures to 

 show the relative proportions of subfractions a and b. 



Freezing point.— This was taken as described on page 553, 

 and is very valuable in indicating the amount of liquid acid 



