10 



* KNOWLEDGE ♦ 



[Jan. 2, 1885. 



made on the lower part of tlie copper, an J, a? we approach 

 the surface of the solution, the proportion of deposited 

 silver gradually decreases. It will be also apparent that a 

 smaller quantity of silver will, under these circumstance.'*, 

 be deposited in a given time by a given battery, and that, 

 in fact, a weaker current is passing than would be the case 

 were the silver precipitated uniformly. This uniformity is 

 accomplished by imparting motion to the copper, and so 

 stirring up the solution. In this way the amount of 

 deposited metal may be increased as much as 15 per cent, 

 and a correspsnding increase in the strength of the current 

 flowing will also be noticed. 



A strong solution, containing only a small proportion of 

 ■water, conducts well, and rapidly produces a rich, silky 

 deposit, but it has the disadvantage of aggravating those 

 troubles to which reference has just been made, and great 

 care is necessary in managing the cell. The cathode 

 becomes covered with a series of variously-coloured deposits, 

 ranging from a good silver tint to lilac, or even to brown. 

 Frequent stirring will, in a great measure, prevent this ; 

 but perhaps the safest way is to weaken the solution 

 by the addition of a little more water, when it becomes 

 much more manageable, albeit a little slower. 



Sometimes it may be observed that the anode is being 

 dissolved much more rapidly than the cathode is being 

 deposited upon. This is a pretty sure indication that an 

 excessive quantity of cyanide of potassium is present, in 

 ■which case the addition of cyanide of silver will, in all pro- 

 bibility, effect all thttt is de.-ired. A little additional 

 water may, however, ahso improve matters ; but dilution to 

 tjo great an extent should be avoided, as the deposit, which 

 naturally forms slowly, assumes an unpleasant dead white 

 ajipearance. The addition of cyanide of silver and cyanide 

 of potassium will soon put matters right in this direction. 



When the requisite amount of silver has been depo.sited, 

 the object is dipped into a cyanide of potassium solution, and 

 then well washed in a stream of water or under a tap, and the 

 washing continued until the whole of the depositing liquid 

 is removed. The object is then dried, and a trough of box- 

 wood or mahogany sawdust having been heated, the object 

 is placed in it and rubbed gently with it to ensure perfect 

 dryness. It is then subjected to the s.-ratchbrush process 

 to rub down the burred surface ja-oduced by the depositing 

 process. Bath-brick apjilied with a stiff hair brush may, 

 however, be substituted for the scratch-brush. The polish- 

 ing (with rottenstone, &c.) and burnishing (with steel or 

 agate burnishers) has then to be gone through. These are 

 processes which are only learned by practice, and which it 

 would repay the student to cultivate and practise frequently. 

 They impart a charm to work which might otherwise look 

 dull and uninteresting. It will sometimes hajipen that, 

 although every precaution may be thought to have been 

 taken, neveitheless some little necessity for cleanliness 

 has been overlooked. The result is, that the apj)lication 

 of the scratch - brush causes the .silver to strip off in 

 such places, and manifestly the work, perfect as it 

 may otherwise be, is efl'ectually spoiled. There is no 

 help for it, save to remove the whole of the silver, make 

 the new exposed surface of the article smooth by rubbing 

 i', with Water of- Ayr stone, and then to replate it. It may 

 also be mentioned here that a similar process may be 

 adopted in order to replate old work from which the silver 

 has been more or less rubbed off in ordinary usage. To 

 remove the silver the article is placed in a stone vessel con- 

 taining some strong sulphuric acid, to which a few crystals 

 of nitrate of potash have lieen added, and which has been 

 raised to a high temperature. The nitiate of potash, or 

 saltpetre as it is more commonly called, dissolves in the 

 acid, and the solution dissolves the silver off the article. 



Care must be taken not to leave the object in the solution 

 too long, otherwise it will stand a good chance of being 

 entirely spoiled. If the silver is not readily dissolved, a 

 little more saltpetre should be added, when the desired 

 efl'ect will be produced. If much of this sort of work is 

 done, the solution becomes exhausted and crystals are 

 deposited on the bottom of the vtssel. As a consider- 

 able quantity of silver is present, an effort should be made 

 to " recover " it. This problem, however, must be defeired 

 for a foitnight. 



CHATS ON 



GEOMETRICAL MEASUREMENT. 



By Richard A. Proctor. 



TUE CONIC SECTIONS. 



(Continued from p. 497.) 



A. We have now to deal with the parabola. 

 J/. Let A S L (Fig. 3) be the axis of the parabola 

 A P K ; S the focus ; K L perp. to A L. We have to 



determine the area A Q K L. Take P Q adjacent points 

 on the curve (which means points pretty near each other) 

 and draw P M, Q N perp. to A L. Now we may con- 

 veniently try the properties of parabolic tangents, because 

 we know that they are related to the lengths of such lines 

 as PM, AM, itc, which are manifestly involved in otir 

 problem. We draw the tangent P YT cutting A Y / the 

 tangent at A (perp. to A L) in Y''. Join S P, S Y, and 

 draw FmJ, Que, Kl perp. to A I, and T d >: perp. to 

 T A L. Then T P Q is a straight line, wlien Q comes near 

 enough to P. Henca the strips P N and P e, being com- 

 ])lementary parallelograms, are equal. But we know that 

 P(/=2P;»:'' hence the strip P« is double the ttrip V n 

 in area. Wherefore the elementary strip PN is equal to 

 twice the elementary strip P oi. Dividing up the whole 

 areas A P K L, A P K / into such areas as P N, Pvi (by 

 taking a multitude of points like P, Q, along A P K), we 

 have each such area as P N in A P K L double the corre- 

 sponding area as P ii. in A P K /. Hence 



the whole area A P K L = 2 the whole area A P K / 

 and parabolic area A P K L=| rectangle A / K L. 



A. That is a simple result. This is, in fact, the first 

 case in which you have shown an area bounded by a 

 curved surface to bear a simple proportion to a definite 

 rectangle. 



JA You must not often expect such a result. 



A. You liave discussed, satisfactorily enough, particular 

 areas taken from the ellipse and the parabola ; but is it not 

 possible to determine the area of any segment of either 

 curve? 



* By a well-known property S Y is perp. to P T, and as P T 

 bisects tlie angle S V d, / S P T = Z tZ P T = Z P T S. Wherefore 

 S r = ST, and PY = YT. Hence Fm = md. But the equality of 

 A T and A M is really a fundamental property of the parabola. 



