Nov. 



1884.] 



♦ KNOWLEDGE ♦ 



381 



neutralise the few diopa of cyanide added for testing pur- 

 poses. 



A little stronger solution of cyanide of potassium is now 

 poured gradually over the precipitated cyanide of silver until 

 the whole of it is again dissolved, a glass stirrer being 

 continually used to facilitate the solution. A small quantity 

 of the cyanide of potassium — say about three-quarters of 

 an ounce — is then added to expedite the subsequent solu- 

 tion of the silver anode. The solution is next diluted with 

 about a gallon of water, and then filtered to remove any 

 impurities that may be present. It is then ready for the 

 bath. 



As hinted in the previous article, the amateur may expe- 

 rience some difficulty in procuring the cyanide of potassium, 

 but a double salt — viz., the ferrocyanide — he should expe- 

 rience no difficulty in getting. With this he may make 

 his own cyanide. The following is the process generally 

 adopted : — Take a quantity of ferrocyanide of potassium, 

 pound it fine, and gently heat it in an iron pan, with con- 

 stant stirring, until quite dry ; treat a quantity of the best 

 carbonate of potash in a similar manner. When they 

 are perfectly dry, add about three parts of the car- 

 bonate to eight parts of the ferrocyanide, and thoroughly 

 mix them ; heat the mixture rapidly in an iron ladle or 

 crucible, until it melts into a clear liquid, when gas will be 

 evolved from its surface. It should be niiintaiued at a 

 moderate or dull-red heat about fifteen or twenty minutes, 

 and until the end of a cold iron rod dipped into it shows a 

 white sample. The fusion should not be continued until 

 the evolution of gas ceases, or the product will be of a grey 

 colour. It should be kept covered as much as possible. 

 By allowing it to stand undisturbed a few minutes at the 

 latter part of the operation, and occasionally tapping the 

 sides of the ladle or crucible, the iron of the ferrocyanide 

 will settle at the bottom as a fine black powder. The 

 colourless cyanide of potassium may then be poured off 

 into a cold iron pan, or upon a thick and cold iron 

 plate. It should be broken up while still warm, and pre- 

 served in a well-stoppered jar. The black sediment, which 

 contains much cyanide of potassium, should be scraped out 

 of the vessel while it is still hot and carefully preserved, as 

 water will at any time dissolve out the cyanide. If the 

 process has been well conducted, the product will be of a 

 clear, white colour, or at most but very slightly grey. A 

 larger proportion of cyanide of potassium is obtained by 

 this process than when ferrocyanide alone is employed, 

 because in the former case one-third of the cyanogen (that 

 which was combined with the iron) combines with the 

 potassium of the carbonate of potassium ; whilst, in the 

 latter case it is lost. The cyanide produced by the fusion 

 of the ferrocyanide of potassium alone is of a greyish-black 

 colour, and is termed " black cyanide.'' 



Except where special preparations are made to produce a 

 deposit having a bright surface, a dead or unpolished sur- 

 face is produced, and if a bright surface is required, 

 mechanical means are resorted to in orJer to obtain it. 

 For some work, however, a dead surface is preferable, and 

 it decidedly has beauties of its own. For such purposes a 

 good solution is made by converting an ounce of silver into 

 the nitrate by the method described in the previous article, 

 and dissolving the crystals in three pints of distilled water. 

 The silver is then precipitated (as the chloride of silver) by 

 the addition of a strong solution of common salt. Having 

 well washed the precipitate, it is next dissolved by the 

 addition of a strong solution of cyanide of potassium, care 

 being taken not to add much more than will dissolve the 

 chloride of silver. The liquid is then filtered through 

 two or three papers, and sufficient distilled water added to 

 make one gallon of solution. 



This solution answers very well, as above-mentioned, for 

 dead work ; more especially is this the case when the solu- 

 tion contains a larger proportion of water than that above- 

 mentioned, when the battery power is weak, and when the 

 anode is small. The deposition of silver under such cir- 

 cumstances would be much slower, and the particles would 

 adhere more firmly to the object being coated. 



A bright surface is imparted to the silver as it is 

 deposited in the following manner : — An ounce of 

 bisulphide of carbon is put into a jiint bottle containing 

 a strong silver solution with more cyanide of potassium 

 than is necessary to dissolve the silver. The bottle, 

 after being well shaken, is stood aside for a few 

 days, and is then ready for use. A few drops of the 

 solution may be poured into the plating bath occasionally, 

 until the surface of the deposited silver attains a sufficient 

 degree of brightness ; care being taken, however, to add 

 the bisulphide solution very sparingly or the depositing 

 solution may be spoiled. This process is most applicable 

 where bright surfaces are required, but which, owing to 

 their irregularity, or to the fragile nature of the object, 

 cannot be burnished in the manner to be presently 

 described. 



CHATS ABOUT GEOMETRICAL 

 MEASUREMENT. 



By IlicH.\RD A. Pkoctor. 



(Continued from page 339.) 



A. I suppose geometrical measurements relating to 

 curves, surfaces, volumes, &c., begin with the lengths of 

 arcsl 



M. We may begin with them, though the simplest 

 methods and the best illustrations relate to surfaces. 



A. How is that] 



J/. You will see if you consider what has to be done in 

 the two cases. Always in measuring curved figures we 

 have to conceive them divided up into great numbers of 

 small parts, these parts being not affected by curvature. 

 Now an arc cannot really be divided into straight lines ; for 

 no parts of it are straight. But we can divide a plane 

 surface bounded by curved lines into great numbers of 

 rectangles, triangles, &c. So we can divide a volume 

 having a curved surface, into prisms, pyramids, parallele- 

 pipeds, itc. A curved surface is as awkward in this respect 

 as an arc, for no part of it is plane. 



A. Can you give me illustrations of your meaning] 



M. Readily. In fact we must start from general illus- 

 trations of these peculiarities. 



Fig. 3. 



A. What then is the difficulty with arcs? 



J/. Suppose A E B (Fig. 3) is an arc we have to measure. 

 We cannot divide this arc into straight portions no matter 

 how small. If we set a number of points C, D, E, F, G, 

 along the arc, we see that the straight lines A C, C D, D E, 

 E F, F G, G B together, come nearer to the true length of 

 the arc than A B. For while these straight lines together 



