230 



THE FARMER'S MAGAZINE. 



made of it, it must be applied in a peculiar mode, and I have 

 contrived a method of using it, which can be employed by any 

 person. This method has been already described in the 

 Transactions of the Society ; but as it appears to have at- 

 tracted comparatively little attention, and bears very much 

 upon the present subject, I trust I may refer to it again now. 

 It is founded on the great chemical principle, that if a 

 sufficient quantity of nitrate of silver be added to common 

 salt, that substance is entirely decomposed and no longer 

 exists in the fluid. This fact can be easily illustrated by re- 

 ference to the experiment already made, in which it was shown 

 howcommon salt could be detected ; for it is only necessary now 

 to add to the solution a further quantity of nitrate of silver, 

 when it is seeu that no further precipitate is formed, the salt ori- 

 ginally present havingbeen decomposed by the previous addition. 

 By reference to well-known chemical law3,whichareuuueceasary 

 for me to explain now, it is found that 29 grains of nitrate of 

 silver are required to decompose one grain of salt. Upon this 

 principle a solution is made, containing 290 grains of nitrate 

 of silver in an imperial pint of water, and its strength is then 

 such that a fluid ounce of it decomposes exactly five grains of 

 salt. The use of this solution for the purpose of determining 

 the purity of nitrate of soda depends on the fact that when of 

 good quality it always contains less than 5 per ctnt. of salt, so 

 that if we take 100 grains of it dissolved in water, and add a 

 measured ounce of the silver solution, we ought to have 

 decomposed the whole of the salt it contained, and none 

 should be left in the fluid ; whereas, if there be more than 

 5 per cent., the excess above that quantity will remain, and be 

 detected by a further addition of nitrate of silver. The mode 

 in which the experiment is made is very simple. One hundred 

 grains of the sample to be examined are weighed in a pair of 

 common apothecaries' scales, and put into a bottle capable of 

 containing about half a pint. Four ounces of warm water 

 (rain water is best) are then poured into the bottle, which 

 is shaken until the nitrate is dissolved. The experimenter 

 then carefully examines the insoluble matter, which should be 

 very small in quantity, not exceeding 1 per cent, aud consisting 

 of grains of sand ; and a little experience will enable him 

 to ascertain by inspection whether the quantity is excessive. 

 An ounce of the nitrate of silver solution is then measured in a 

 common glass measure, and poured into the bottle, which 

 is shaken violently for some time, and then allowed to stand 

 at rest for five or ten minutes, when the white curdy matter 

 deposits aud leaves the supernatant fluid nearly or altogether 

 clear. When this is the case, a few drops of nitrate of silver 

 are added, and if the fluid remains clear and shows no further 

 milkiness, then the sample must have contained leas than 

 5 per cent, of salt, and is of excellent quality. But should it 

 become muddy, another ouuce measure of the nitrate of silver 

 is added, the bottle shaken, allowed to stand until clear, and 

 the clear fluid again tested with a few drops of the 

 nitrate, and if it remains clear, then the sample contains 

 more than 5 and less than 10 per cent, of salt. Should it 

 still give a precipitate, the same manipulations are gone 

 through again as often as may be requisite, until the clear 

 fluid cease to become muddy. Every ounce of nitrate of 

 silver which has thus been used corresponds to 5 per cent, 

 of salt, so that if three ounces had been added there must be 

 from 10 to 15 percent, of impurity and so on over. In this 

 way the experimenter can always ascertain within 5 per cent, 

 the amount of impurity. But it would probably not be advi- 

 sable for him to go beyond the first additions ; and should he 

 find that there is more than 5 percent., then he should suspect 

 that his sample is inferior, and have it carefully analyzed. It 

 IS important to notice that, by this method of testing, we pro- 

 ceed by steps of 5 per cent., and that, for instance, no distinc- 

 tion will be made between a sample containing 6 and another 

 containing 9 per cent, of salt ; but it is obvious that any degree 

 of precision may be arrived at, provided we add the solution of 

 nitrate of silver, not by whole ounces, but by smaller quanti- 

 ties, for every fifth of an ounce must correspond to 1 per cent. 

 The objection to using it in this way, however, lies in the delay 

 consequent in waiting between each addition ; but it is very 

 easy to make an exact experiment for the purpose of checking 

 the correspondence between the article delivered and the 

 analysis by which is was sold. Thus, supposing the analysis to 

 bear that the sample contains 3'5 per cent, of salt, it would 

 only be necessary to add three-quarters of an ounce of the 

 nitrate of silver ; and if that precipitated the whole of the salt, 



then the result is satisfactory. It is, in fact, only necessary to 

 take a little more of the nitrate than is requisite to precipitate 

 the whole of the common salt or chloride of sodium stated in 

 the analysis, and to ascertain whether it has actually done so 

 by observins; whether the solution remains clear when a few 

 drops more of the nitrate are added. I have enlarged thus 

 much on the detection of common salt, because it is the chief 

 adulteration of nitrate of soda, against which it is necessary to 

 guard. I have already alluded to sulphate of soda as a possible 

 adulteration, although I have never seen it, and suspect it 

 could scarcely be used without exciting suspicion. It would 

 be possible, however, to detect it by a process similar to that 

 used for salt, employing a solution of 171 grains of the sub- 

 stance called chloride of varium in a pint of water ; but it is 

 not so convenient, because the precipitate formed does not 

 deposit so rapidly, and it would be necessary to leave 

 the fluid for ten or twelve hours between each addition. 

 Nitrate of potash may be tested by means of the same 

 solution of nitrate of silver. As that sold for agricultural 

 use is generall}' very inferior, and often contains large quan- 

 tities of sulphates and other salts, great care must be exer- 

 cised during the process. We now come to the most 

 important and extensively employed of all manures, namely, 

 guano ; and with it we meet with increased difficulties, for 

 we pass from definite substances, such as those we have 

 just treated of, to another which is far from uniform ; for 

 we have guano or all kinds, from Peruvian, containing 16.5 

 per cent, of ammonia, and 22 to 25 per cent, of phosphates, 

 to Saldanha Baj', containing only 1.5 per cent, of the 

 former and above 60 of the latter, and all genuine, that is 

 to say, all exactly as they are imported, though, of course, 

 varying very widely in quality. We lose the definite 

 standard which is possessed by the simple cnmpounds, and 

 hence the testing a guano must of necessity be less definite; 

 and we are forced to rely upon a variety of different obser- 

 vations, from the combination of which a probable conclu- 

 sion can be drawn. These observations apply with less 

 force to Peruvian guano, which, though variable in qualitj', 

 within pretty wide limits, cannot be adulterated without 

 prodixcing such alterations in its composition as can be 

 very readily detected ; and hence, when deciding upon the 

 quality of a guano, it is necessary to draw a short line of 

 demarcation between it and all other varieties. In exa- 

 mining a sample of Peruvian guano, it is necessary first to 

 attend to the colour, which should be pale. If it is dark, 

 the sample is probably damp ; and if, when a handful is 

 strongly squeezed, it coheres into a mass, this is still more 

 probable. It should contain lumps, having a rather paler 

 colour than the mass of the guano, when they are broken 

 across ; and sometimes they are white, and have a crystal- 

 lized texture. The entire absence of lumps is always a 

 suspicious circumstance ; but adulterators are now so well 

 aware of the importance attached tft them, that great care 

 is taken not to break them down, and their presence is 

 therefore no guarantee that the guano is genuine. The 

 observer will next fill a bushel measure with the guano, 

 and weigh it ; and if the weight exceeds 721bs. the guano is 

 probably adulterated. But by far the most conclusive test 

 of adulteration is one which approaches to an analytical 

 process, although it is one which can be performed without 

 difficulty. It depends upon the fact that in a genuine 

 Peruvian guano the moisture and organic matter make up two- 

 thirds of its whole weight ; and as these are volatile at a red 

 heat, it follows that when burnt it should leave an ash not 

 exceeding one-third of its weight. For this experiment a 

 small porcelain cup or capsule is required ; and after this has 

 been accurately counterpoised, by means of sand or any other 

 matter, 20 grains of the guano are weighed into it. The 

 capsule is then placed in a hot place by the side of a fire, aud 

 when it has become hot it is carefully placed upon the red-hot 

 coals on a clear fire, and left there for fifteen or twenty mi- 

 nutes, and is then removed and allowed to cool as gradually as 

 possible, so as to avoid cracking the capsule, and weighed 

 again. The residue ought to weigh about 7 grains ; and if it 

 exceeds 8 the guano is certainly adulterated. The value of 

 this test depends upon the fact that all the substances used 

 to adulterate guano — such as sand, sulphate of lime, chalk, 

 &c., &c. — are fixed in the fire, and as they are generally used 

 very liberally, the chances are that the quantity of residue is 

 materially increased — in point of fact, it will probably range 



