This paper gives a procedure for the quantitative determination of hydrocarbons in sediments, discusses the problems of making these measurements, and presents precision data for the procedure recommended. The standard deviation of making a single hydrocarbon measurement in ancient sediments was found to be a function of concentration of the hydrocarbon, for example s = 3, 10, and 400 ppm at 10, 200, and 5,000 ppm, respectively.

The effects of several variables on the results of hydrocarbon determination, each within the generally used operations of extraction, isolation, and chromatography, are as follows.

Extraction:
Samples of 10 micron particle size gave up twice as much extractable material as samples of 30 micron particle size. This material included additional hydrocarbons. No additional hydrocarbons were extracted from 5 micron particles, compared with 10 micron particles, though 5 micron particles gave 2.75 times more extractables than 30 micron particles. Using a specially designed apparatus, extraction was 90 per cent complete in 10 hours and 99 per cent in 20 hours. Benzene is preferred over mixed solvents for extraction of hydrocarbons.

Isolation:
Evaporation of benzene from 10 to 150 milligram portions of residue having approximately 200°C. initial boiling point is performed with considerable uncertainty because of sample volatility. This is a major factor presently limiting precision and accuracy of the determination.

Chromatography:
Extract residues usually are resolved on silica gel having 32 A pores only if the proportion of sample is less than 0.1 per cent of the adsorbent. Using a layer of 170 A gel over the 32 A gel increases the adsorbent capacity to 2 per cent.