Considerable uncertainty still exists regarding the detailed arrangement in protein‐lipid molecular associations found in serum lipoproteins, plasma‐membranes of cell organelles, the myelin sheath of nerve, and many other structures of paramount importance to biological mechanisms both in health and disease. Working hypotheses concerning such structures must eventually be tested by comparing known properties with those suggested by exact stereomodels. In this type of study, orthogonal projections of the molecules offer several advantages over the tri‐dimensional stereomodels from which they can be derived by a described photographic process. The rules governing the configuration to be given the molecular models used for this purpose are discussed, and the possible applications of the resulting diagrams are described and illustrated by numerous examples taken in the lipid field. These rules are then applied to the lipids in the myelin sheath of nerve. Striking similarities in the configurational features of the two main classes of myelin lipids, the phosphatidyl (dal) and the sphingolipids, immediately suggest very similar models for them. On the other hand, numerous independent observations have indicated the highly probable occurrence of bimolecular complexes involving members of either class with cholesterol. Such complexes are exemplified by proposed models of cholesterol‐lecithin and cholesterol‐sphingomyelin units. It is demonstrated that the cohesional forces at play should indeed promote stable complexes of this kind. All configurational features of the lipids in myelin fit these basic models. Dimensional variations induced by a broad spectrum of component fatty acids, affect only the length of the resulting complex units. Moreover, the tail‐to‐tail arrangement of these units provides paired elements of the same length. The latter corresponds exactly to the fundamental dimension predicted for the bimolecular leaflet by low angle X‐ray diffraction studies on fresh, unfixed myelin. A model of the bimolecular lipid leaflet produced by parallel grouping of the paired elements is discussed.

Ozonolysis of oleic acid in methanol, a reacting solvent, with s~i b s e q~~e n t decomposition of the ozonide products by hydrogen peroxide in formic acid, gives yields of principal dicarboxylic acid fission products exceeding 95y0 with a mini111~1m of secondary acidic products.The method is highly reproducible and offers unique advantages in the total recovery of the dicarboxylic acids and the eliininatioil of peroxidic materials. The principal non-acidic byp r o d~~c t s were tentatively identilied as the CS alcohols and their for111yl esters.

No abstract

All melting points were determined by means of the instrument, A fresh sample of the material was used for each determination; in the case of sulfanilamide, repeated determinations were also made automatically on a single sample

At the time the study was undertaken, no method existed for effectively separating monocarboxylic acids from C2 to Ci2 and dicarboxylic acids from C, to Cio. Results described in the literature for monocarboxylic acids were difficult to reproduce, some of the methods were laborious, and factors affecting the resolving power of the column were not well defined. A new multicolumn technique allows the separation of monocarboxylic acids from C2 to Ci2, and with only a modification of the column mixture composition, dicarboxylic acids from C, to Cio.Recoveries are often better than 95%. An apparatus well adapted to partition chromatography in general,