We tried thus to utilize the products resulting from the breaking down of body constituents or of other materials in order to induce through their administration, the appearance of the second defense mechanism. Applying the dualistic concept, we separated thus in the products of hydrolysis of different materials, those with an acid character from the group with basic and alcoholic characters. Various materials were thus hydrolyzed using KOH, NaOH or ammonia. The soluble part, separated, was treated with an acid and a precipitate obtained. After washing it, this precipitate was redis solved by alkalizing to a pH still below neutrality. This has represented the "acid fraction." Besides acid lipids, this fraction contains also acid protein groups and even humic acids.

The part which remained insoluble after treatment with KOH (separated from the soluble part) was treated with an acid. The part which became soluble was then separated, reprecipitated by alkali, and partially redissolved by bringing the pH, through acidification, near 7. This represents the "alkaline fraction." With different degrees of chemical hydrolysis, various fractions—more or less broken down—are obtained for both the acid and alkaline fractions. The degree of this "digestion" has appeared highly important. The amount of the products obtained decreases for an insufficient hydrolysis as well as for a too highly pushed hydrolysis.

According to the mechanism mentioned above it was expected that these fractions, resulting from the breaking down of body constituents or of the antigen and corresponding to the effect of the first enzymatic defense mechanism, would induce the second step of the defense mechanism. This would correspond in part to the intervention of the properdin system and of the lipidic defense. It has as characteristic the fact that it would appear only within a certain time. The following experiment illustrates this clearly. The "acid fractions" of human blood, hydrolyzed by KOH was obtained and then injected intraperitoneally to mice. At different intervals following this injection the mice were inoculated with 3,000,000 microbes of a fresh culture of Bac. proteus. In controls this inoculation would result in a 100% lethal infection. No protection was seen to appear in the 16 hours following the injection of the "acid fraction." At the 16th hour, 1/20 were protected. This protection increased with time to be complete after 22 hours, when all the animals survived. This protection was still present after a few days.

The inoculation of the nontreated blood in the same proportion was seen incapable of conferring the same degree of defense, a fact which indicates the importance of the breaking down process in this "24 hours" defense. These results are similar to those obtained by I. A. Parfentjev with malucidin, a product of hydrolysis of yeast.

Another application of the same concept was in the use of the lipido proteic complexes.

In a group of research studies, we utilized the products resulting from the bond between an antigen and a lipid, with the intent to obtain a lipido proteic antigen and through it, a lipido proteic defense response. Often the mixture of the antigen with the lipidic preparations appeared sufficient.

Fatty acids, such as oleic, linoleic, arachidonic or eleostearic, acting directly upon the killed typhoid microbes were usually seen to enhance the production of agglutinins and of specific immune antibodies. The same effect was produced by lipoacids of the same species as the test animal. Lipoacids of guinea pigs were especially active in promoting the appearance of agglutinins but less potent in inducing the appearance of immune antibodies. The lipoacid fraction of bacteria such as B. subtilis, coli, diphtheria, acting in vitro upon typhoid killed microbes, led to the appearance of antibodies against typhoid microbes but produced almost no antibodies against those microbes from which the fatty acids were obtained. The lipoacid fraction of tubercle bacilli bound to killed typhoid microbes was seen to induce agglutinins but seemed to reduce and even prevent the appearance of immune antibodies. The same influence was seen with the lipids obtained from the seeds of Bixa orellana but was less accentuated for the lipids from fish and squid. Butanol and especially heptanol were seen to retard the appearance of all antibodies, allergic and immune.