Films of oil which have originally covered the whole of the surface of an area of water frequently break up and leave the oil only in patches, and the rest of the surface uncovered. An investigation has revealed that this may be the result of either or both of two circumstances.

It has been shown that instability may result from :—

I. Certain relationships between the volatility of the aromatic and aliphatic hydrocarbons composing the oil mixture.

II. The presence of certain polar substances dissolved in the oil.

I. The volatility relationships which determine the stability or instability are as follows :—

(a) If the aromatic hydrocarbons are more volatile than the aliphatic, a tendency exists for the oil to become concentrated into one central thick lens with the rest of the surface uncovered by oil.

(b) If the whole of the aromatics are more volatile than the whole of the aliphatics, this action is very rapid and occurs whatever the proportion of the mixture.

(c) If the difference in the volatility is only small, the phenomenon is only observed when the proportion of aromatics to aliphatics is very near to unity.

(d) If the boiling range of the aromatics overlaps the boiling range of the aliphatics (as occurs when two oils containing both types of constituents are mixed), the phenomenon is much less clearly defined.

(e) Olefinic oils, e.g., oils obtained by cracking, behave like aliphatic oils in so far as this behaviour is concerned.

It is believed that these changes would not occur if the oil was extremely pure, since lenses are only stable if subject to lateral pressure from an invisible film of contaminating material.

II.—The effect of polar materials can be summarised:—

(a) Solutions of certain organic substances in oil tend to make unstable films on the surface of water; rupture of such a film (by wind or other agitation) allows a unimolecular film of the added substance to spread at the air-water interface. This forces back the oil film. The greater the agitation the less the area of surface eventually covered by oil. In this class of substances are fats, fatty acids, naphthenic acid (a natural constituent of petroleum), some sulphonic acids and cholesterol.

(b) Certain other substances act in a contrary manner. They do not promote instability but actually counteract and, if present in sufficient proportion, overcome the instability produced by the first type of substance. The substances falling in this group are all at present of unknown composition, but include “ cracked spirit gum ” (a material derived by polymerisation of olefinic hydrocarbons), certain resins, and some substance present in petroleum and remaining in the residue after distillation. The last of these acts as a natural “ stabiliser ” and masks the instability phenomenon to some extent in commercial oils. It is found in largest quantity in the undistilled residue of aromatic concentrates.

(c) The quantitative relationship between the “ stabiliser ” and the substances promoting instability is not linear. Successive increments of “ stabiliser ” equalise larger and larger amounts of the other.

(d) Long-chain molecules such as fatty acids are the most difficult to overcome by “ stabiliser.” Naphthenic acid is more easily counteracted, cholesterol still more easily.

The following recommendations can therefore be made as to the desirable composition of anti-malarial mixtures, so far as film stability is concerned :—

1. The oil should contain either a small aromatic content or a very high one ; it should not contain 50 per cent. aromatics unless they are very high-boiling (e.g., of lubricating base fraction).

2. It should consist of a mixture of wide and overlapping cuts of oil.

3. It should not contain fats or fatty acids added as spread-aiders. If spread-aiders are desired, resins should be employed.

This research was financed in its entirety by a grant contributed equally by the Anglo-Iranian Oil Company, the Asiatic Petroleum Company, and the Burmah Oil Company. To the research chemical staffs of these companies I express my thanks for samples of oil and for much technical information and advice. My thanks are also due to Professor P. A. Buxton and Professor N. K. Adam for advice and help.