1. Chemical mutagens have been shown to act on gene loci which are apparently stable to X-radiation. By the use of a group of alkylating compounds it was possible to induce nearly 200 ‘new’ sex-linked recessive visibles (ranks 1 and 2), different in phenotype and genetic position from those induced by the physical agent. 2. The ratio of sex-linked recessive visibles/lethals in the same sample of treated chromosornes is exceptionally high for a particular amino-acid mustard (p-N-di(chloroethyl)phenylalanine); which mutates 2–3 times as many morphogenesis loci (relative to lethals) as X-rays or any other alkylating compound. 3. The distribution of the loci of recessive lethals along the X-chromosome is the same for three representative alkylating mutagens, and is significantly different from that for X-rays. A significant difference also occurs in the distribution of theF 1 viable breaks induced by an imine (tri(ethyleneimino)triazine) as compared to X-radiation. 4. The rate of induction of small deficiencies by the alkylating compounds is nearly twice that produced by mutagenically equivalent doses of X-rays. Evidence is available that chemically induced deficiencies result from errors in gene reproduction rather than chromosome breaks. 5. There is a shortage in major structural rearrangements in viable sperm, as well as among the sex-linked recessive lethals, under the effect of the alkylating agents compared with mutagenically equivalent doses of X-rays. This shortage, in the case of the tri(ethyleneimino)triazine rearrangement lethals, has been shown to be due to differences in the properties of the chromosome breaks induced by the imine and X-rays, rather than a lower efficiency in the induction of the primary chromosome breaks by the chemical agent. 6. The dose effects as regards the induction of sex-linked recessive lethals, dominant lethals and viable chromosome breaks, are identical for tri(ethyleneirnmo)triazne and X-rays. This suggests that the hypothesis of direct action of radiations on the genetic material is not conclusively proved by evidence based solely on dose relations. 7. Among the sex-linked recessive lethals induced by various closes of tri(ethyleneimiao)triazine there is a fixed proportion showing cytoiogically detectable chromosome aberrations: major rearrangements and deficiencies. Within this fixed proportion, the fraction with rearrangements increases and that with deficiencies decreases with increase in dose. The bearing of this observation on the mechanism of induction of chromosome breaks by chemical means has been outlined.