556 University of California Publications in Agricultural Sciences [Vol. 1 



To all this, however, there must be added some other con- 

 siderations. One of them serves to qualify, in some measure at 

 least, tlie remarks made above, and the other to supplement them. 

 While the metals studied by us do not seem to have given evi- 

 dence, under the conditions of our experiment, of any serious 

 injury to barley, a non-metal, arsenic, has given marked evidences 

 of toxicity to barley under similar conditions. Arsenic, being 

 found frequently in conjunction with the other elements in the 

 vicinity of smelters, is necessarily a subject worthy of attention. 

 Our results with its use in soil cultures are not yet ready to 

 be reported, but we hope sometime in the near future to publish 

 them. Suffice it to say now that such compounds of arsenic as 

 arsenic trisulfide and Paris green have proved to be extremely 

 toxic to barley in both heavy and light soils, while lead arsenate 

 has proved to be only slightly toxic. Whether or not arsenic 

 oxide, w^hich is the form to be expected in lands in the vicinity 

 of smelters, will act similarly remains to be shown by further 

 experiments which are now being planned by us. 



We are constrained to add to the foregoing that we have 

 borne in mind the difference in the effects produced on a toxic 

 material by the change in a soil's constitution. Indeed, our 

 experiments with copper in three widely different types of soil 

 testify to that fact ; and while we have found marked differences 

 in the degrees of stimulation and toxicity of copper in the differ- 

 ent soil types, all of the latter appear to have given both stimu- 

 lation and toxicity. Even in the sandy soil in which the toxicity 

 of CuSO^ became manifest at the lowest concentration for any 

 of the types of soil studied, as much as 0.03 per cent of CuSO^ 

 of the dry weight of the soil still acted as a stimulant to barley. 

 Considering that CuSO^ is an easily water-soluble salt, it would 

 be reasonable to expect that such compounds as Cu(0H)o.CuC03, 

 which are the usual forms to be expected in soils near smelters, 

 could be tolerated by plants in much larger quantities. 



If, as appears to us reasonable, we should be able to accept 

 the data above offered by us, at least as tentative evidence that 

 we have little to fear from the solids of smelter wastes in the 

 contamination of our irrigation water-supply and therefore in 



