ANALYSIS OF PLANT*. 35 



the impure ash is weighed, and all the then soluble matters dissolved out, when the residue 

 is again burned. This method is adopted only when the ash fuses at a very low tempera- 

 ture, or when the proportion of mixed alkalies is comparatively large. Generally the ash 

 has been used soon after it has been prepared c if it stand it is dried previous to analysis. 



II. The analysis begins with the weighing two or three parcels of 20 grs. each : some- 

 times a less quantity is used. One parcel is used for the earth, alkalies and phosphates ; 

 another for chlorine, sulphuric acid and organic matter. I obtain the chlorine from one 

 half, and the sulphuric acid and organic matter from the other half. The other parcel is 

 dissolved in hot hydrochloric acid. In many cases, this is a speedy operation : time, how- 

 ever, is given in the straw and chaff of the cereals, when silicates may be expected. We 

 usually obtain in these cases a gelatinous sokition, which indicates the perfect action of 

 the acid. Silica is obtained by filtration : it is ignited and weighed. The filtrate contains 

 the phosphates, lime, magnesia, potash and soda. The phosphates are thrown down by 

 excess of ammonia ; the lime, afterwards, by oxalate of ammonia. The filtrate, after the 

 lime has been separated, is divided into two equal parts : one is used for magnesia; the 

 other, for the potash and soda. To obtain the magnesia, I prefer the phosphate of soda 

 with ammonia. We have used the peroxide of mereury, and it saves time and some labor-; 

 but we have been better satisfied with the phosphate of soda. The half reserved for potash 

 is evaporated, and the salt exposed in a Berlin capsule to low ignition. It is then weighed, 

 and the result is set down as chlorides. These are dissolved out with water, and filtered ; 

 the filtrate is evaporated again, and the dry residue submitted to the action of absolute 

 alcohol. The insoluble matter remaining in the filter, which is washed with alcohol, is 

 dried, and heated rather strongly, and weighed and set down as chloride of potassium, 

 from which the amount of potash is calculated. The chloride of potassium and magnesia 

 being subtracted from the chlorides, gives the chloiide of sodium, from which the soda is 

 calculated. The insoluble matter in water is also noted. This method, though not re- 

 commended in Fhesenius's work, I can not but regard as quite accurate. Fresenius 

 says that chloride of poteissiura is nearly insoluble in absolute alcohol ; and we have found, 

 on testing the alcohol made in the laboratory, that it does not dissolve a perceptible quan- 

 tity of pure chloride of potassium, while chloride of sodium dissolves in it rather freely, 

 and chloride of magnesia at once. 



III. After going through with the preceding work, the phosphates which have been 

 preserved are taken up for analysis. They are redissolved in hydrochloric acid, and fil- 

 tered to free them from soluble silica. The filtrate from the soluble silica is mixed with 

 ammonia in slight excess ; and while the mass is still wet, acetic acid is added, which 

 dissolves all but the phosphate of peroxide of iron : this is separated by filtering, and dried, 

 ignited and weighed. From the filtrate, oxalic acid throws down the lime, which is also 

 filtered, ignited and weighed. The magnesia is now thrown down by phospiiate of soda 

 and ammonia, the latter in slight excess, or else the acetic acid will still hold the magnesia 

 in solution. 



I can not hut regard this method of treating the phosphate, as preferable to that where 



