some varieties of' South American Guano. 127 



actual mixture of soda and potash salts, and the results ap- 

 proximated very closely to the truth; and upon the whole, 

 weighing the advantages of this indirect system of Griffin's 

 with the positive modes of estimating the potash, either by 

 perchloric acid, tartaric acid or platina salt, 1 believe this 

 indirect mode to afford, in careful hands, at least as accurate 

 returns, if not more so, than the direct modes of estimating 

 the potash, and ascertaining the soda from the loss. I may be 

 wrong in attributing this process to Mr. Griffin, as Barresvvil 

 and Sobrero assign it to M. Mayer, with a date anterior to 

 Griffin's description in the Philosophical Magazine. 



With respect to the neutral phosphate of lime stated to 

 exist in the sample of " Concrete Guano, No. 6," and also in 

 No. 2, I may state that it precipitated as neutral phosphate ; 

 and if I had estimated it as bone-phosphate, the sum of the 

 analysis would have fallen very far short of the quantity em- 

 ployed ; whilst by estimating it as I have done, the loss was most 

 trivial, amounting only to a common error of analysis. Of the 

 solubility of the phosphate of lime, and of the ammoniacal 

 magnesian phosphate in boiling water, it is simply necessary 

 to refer to the already established solubility of these salts in 

 neutral solutions containing organic animal matter, and also 

 to state that these solutions were invariably faintly acid to 

 litmus paper. 



I was led to suspect the existence of a volatile acid in guano, 

 but I in vain tried to isolate it, if indeed it exists, and the acid 

 odour and reaction are not truly to be attributed to the re- 

 action of the sulphuric or phosphoric acids, the agents made 

 use of in the attempt, upon the organic portions of the guano. 



The quantity submitted to analysis was in each case either 

 300 or 500 grs., and the results were calculated to 1000 parts 

 for the convenience of ascertaining the per-centage, and for 

 greater readiness of comparison. 



The water was determined approximately, by drying till the 

 portion lost no more weight, without any sublimation taking 

 place. Water and ammonia were volatilized. This estima- 

 tion of the water was corrected by subtracting the sum of 

 the weight of the soluble salts, and portions insoluble in cold 

 water, ascertained during the subsequent process, from the 

 original weight of guano employed. A given weight of the 

 sample of guano to be examined was ground up with cold 

 distilled water; when the insoluble portions had subsided, the 

 solution was passed through a filter and the residue washed 

 with two or three successive portions of distilled water; the 

 insoluble residue was then placed on the filter, and treated with 

 distilled water until salts of lime and barytes, being added to a 



