626 Prof. T. Godlewski on the Action of 



4. Influence of Positive Colloids. 



The positive colloid employed was a dyalised solution of 

 ferric oxide ; it was the ordinary commercial article, con- 

 taining 6*62 gr. of ferric oxide per* 100 cm. 3 The experiments 

 were conducted on the same lines as before ; their results are 

 plotted in fig. 3. 



It will be seen that the influence of colloidal ferric oxide 

 is quite special. As the theory of colloids might lead us to 

 expect, the amount of products ("Eta A and RaC) deposited at 

 the anode decreases at first : thus giving evidence of the 

 precipitation of oppositely electrified hydrosols. In certain 

 experiments a slight diminution of the quantity of RaC 

 deposited at the cathode was noticeable in the low concen- 



trations, while the relative amount , / "X increased to a small 



RaC 



extent. This result may perhaps be ascribed to a greater 



adsorption of RaC by ferric oxide than by the remaining 



atoms of RaB. 



In concentration y = 6*6.10 -3 mg. we observe a very 

 great falling off in the activity of the anode ; with a higher 

 concentration the activity of the anode is exceedingly slight. 

 On the other hand, from the concentration y = 6'6. 10~ 2 mg. 

 upwards the appearance of RaA and RaC at the cathode is 

 manifest, so that all the products are deposited at this elec- 

 trode. In these concentrations, then, the products which 

 usually form negative hydrosols become positively electrified 

 and are deposited at the cathode. 



The conclusive method of proving the adsorption of nega- 

 tive hydrosols is to add ferric oxide in considerable quantity 

 to water freshly saturated with emanation and to proceed 

 with the electrolysis as rapidly as possible. The deposit at 

 the cathode in this case will be found to consist of almost 

 pure RaA. 



It may be added that the results adduced in the foregoing- 

 pages have been found in each case by taking average values 

 out of a large series of individual measurements. Though 

 the sequence of the curves (corresponding to various con- 

 centrations) in different experiments was always the same,, 

 the particular results were observed to be subject to soma 

 slight variations, especially at lower concentrations. Subse- 

 quently it was found that slight alteration of the circum- 

 stances of the experiment (as for instance a change in the 

 rapidity of mixing, variation of the rate at which the colloid 

 was supplied, and even the concentration at which a given 

 quantity of colloid was added) — all this exercises a consi- 

 derable effect on the result obtained. In order to arrive 



