1. A technique for comparing the spermicidal powers of pure substances is described in detail. It is hoped that this may be accepted as the standard technique for the purpose.

2. The killing concentration of each substance is determined by this technique. The killing concentration is defined as the lowest concentration, in the series 2, 1, ½, ¼ per cent., etc., which suffices to kill every guinea-pig sperm suspended in glucose-saline solution in half an hour at the temperature of the body in four consecutive experiments, the majority of the control sperms being moderately or very active.

3. 36 substances have been graded by this technique.

4. Mercuric chloride and formaldehyde were found to be the most spermicidal substances. The killing concentration of each per cent.

5. Hexyl resorcin kills at per cent., soaps at per cent.

6. Formaldehyde and hexyl resorcin, among other substances, seem likely to be useful as contraceptives.

7. The significance of the high spermicidal power of soaps is discussed.

8. Quinine bisulphate and chinosol, which are perhaps more commonly used as contraceptives than any other substances, only kill at ½ per cent.

9. Certain very poisonous substances have very slight spermicidal powers. This applies to potassium cyanide, prussic acid and strychnine hydrochloride.

10. Foaming mixtures, consisting of acids and sodium bicarbonate, could probably be used alone as contraceptives.

11. The acrosome is the part of the sperm most vulnerable to spermicides. It tends to swell up and burst.

12. Sperms are very susceptible to changes in osmotic pressure.

13. It is the hydrogen ions and not the anions of acids that kill sperms.

14. The suggestion that tribasic acids would be found to be more spermicidal than dibasic, and dibasic than monobasic, is not substantiated.

15. Substances which reduce surface tension are often effective spermicides.

16. The characters of the ideal chemical contraceptive are discussed.

17. The need for co-operation in research in chemical contraception is stressed.