H6 



J. E. MAYER. ESQ. ON THE WATER 



results of filtration by mixed agency, as shewn by Nos. 4 ? 

 8 & 9, shew not only great diversity of results, but that 

 those obtained by means of sand and charcoal. (No. 4) 

 require, in order to make such materials effective, certain 

 other precautions which could not be adopted in this or that 

 case, without vitiating the first principal of comparative 

 investigation, — that, viz., of placing all the substances expe- 

 rimented on under " precisely similar conditions," — and it 

 will be observed that in none of these experiments has this 

 canon been departed from. The simplest means, the same 

 sized funnels and the like quantities of materials were in 

 each instance employed. In no instance have special 

 arrangements, packing, or adjusted pressure been resorted 

 to ; but without these means, the filtration by double 

 agencies, as shewn by Nos. 8 and 9, appears to exceed 

 in a slight degree the purification that is obtainable by 

 single mechanical agencies. Whether the magnetic oxide 

 of iron acts by its insolubility and specific gravity in such 

 a manner as to form very minute pores through its 

 mass, — or by any special mode of action, has not been 

 investigated by us. The artificial product is a finer powder 

 than the natural magnetic oxide, and it acts both singly 

 and in conjunction with Swedish filter paper better than 

 the natural product, which considerations tend to impart 

 the idea that, after all, it acts mechanically, and not 

 from its possessing any peculiar purifying power. From 

 a partial examination of Spencer's Filtering Apparatus, it 

 was evident, 1st, that a sponge under a graduated pressure 

 was employed as a preliminary mechanical means of inter- 

 cepting impurity, 2ndl} T ,that pieces of sponge cut very small, 

 were interspersed through the material in the body of the 

 filter. "What further special arrangements may have been 

 employed we are unable to say, — as it was not thought worth 

 while to destroy the Apparatus for the sake of such informa- 



