The above method is inapplicable when the proportion of essential oil is very small, as in flavoured beverages and some medicines. In such cases, it is often practicable to make an approximate determination by distilling the liquid and extracting the distillate, sufficiently diluted, with chloroform or carbon tetrachloride. The iodine absorption of the oil thus extracted is then determined by adding Hubl's or Wijs's iodine solution to the chloroform or carbon tetrachloride extract, and proceeding as usual. A comparison experiment is made at the same time with a solution prepared to imitate the sample under examination, but containing a known quantity of the oil in question. This solution is distilled, and treated side by side with the sample, so that the operations are similar in all respects. From the results the approximate quantity of essential oil can be calculated. It bears the same ratio to the known quantity in the comparison sample as the corresponding iodine absorptions have to one another. (This is strictly correct only if the iodine "values" of the two specimens of oil are the same; but for the small quantities here concerned the variations are of no moment.)

Generally, 25 or 50 c.c. of the sample may be diluted and distilled till 75 or 100 c.c. have passed over, the distillate extracted three times, with 12, 8, and 5 c.c. of chloroform or carbon tetrachloride (25 altogether), and 5 or 10 c.c. of Hubl's or Wijs's iodine solution added to the separated extract. A "blank" experiment on a similar quantity (25 c.c.) of the solvent is of course made in the usual manner.

Fairly good determinations of small quantities of essential oils can also be made in many cases by saturating the distillate with sodium chloride, and extracting the oil with petroleum ether, as in Thorpe and Holmes's method for purifying alcohol. The petroleum ether is allowed to evaporate at the ordinary temperature in a tared beaker, and the residual oil weighed. (See also the methods described under "Liqueurs.")