222 SELECTION AND TJSE 



be laid orer an old one, the old layer will be softened, and 

 if there should be any tendency to a vacuum in the cell, 

 4w softened cement will be unable to resist the outside 

 <v *essure, and will creep in and spoil the object. So, too, 

 with varnishes or cements formed chiefly of drying oil or gold 

 size. If the different coats be laid on too thickly or too 

 rapidly, the part that is beneath cannot easily harden, but will 

 remain for a very long time in a semi-liquid condition. We 

 have just removed some brass rings from slides to which they 

 were attached four months ago by means of gold size, and 

 although the outer surface of the cement was hard and dry, the 

 interior was quite liquid, freely soiling the fingens. 



GOLD .SIZE. The most extraordinary recipes have been given 

 for the preparation of this cement, which is in reality nothing 

 but good linseed oil rendered very drying by the usual 

 methods. Gilders frequently make it into a semi-paint by 

 adding coloring matter, thus forming a ground of a shade 

 similar to the gold they use, and this seems to have misled some 

 of our best writers. There is no ochre, litharge, or anything 

 of the kind present in good gold size. It does not pay to pre- 

 pare gold size in small quantities, and it may be obtained from 

 any color dealer. The older it is the better, and it is well, 

 therefore, to lay in a good stock, which must be kept carefully 

 corked. The wqrking supply should be kept in a small bottle. 

 This is the favorite cell making material employed by Dr, Car- 

 penter, and it is certainly the most reliable cement we have. 

 It adheres firmly to glass, and if laid on in very thin successive 

 layers, tolerably deep and very durable cells may be built 

 up, but the process requires considerable time, otherwise the 

 under layers will remain soft. It has this great advantage over 

 asphalt, damar, and other cements composed of solid materials 

 dissolved in some menstruum, that fresh coats have but very 

 slight action on the old layers on which they may be laid. It 

 mixes with turpentine, and consequently with most materials 

 soluble in turpentine, but when once dry and hard, turpentine, 

 alcohol, ether, etc., have little or no action on it. It does not 

 mix with alcohol, and therefore cannot be mixed with the 

 solution of shellac in alcohol in any of its forms. 



