Surely in face of these facts we must admit that cement joints can be made with fair success. I do not know the composition of M. Regnault's mastic, but Faraday (Manipulations, Sec. 1123) gives the following receipt for a cement for joining ferules to retorts, etc: Resin 5 parts. Beeswax 1 part. Red ochre or Venetian red, finely powdered and sifted 1 part. I believe this to be substantially the same as Regnault's mastic, though I have never analysed the latter. For chemical work the possibility of evolution of gas from such a cement must be taken into account, and I should certainly not trust it for this reason in vacuum tube work, where the purity of the confined gas could come in question. Otherwise it is an excellent cement, and does not in my experience tend to crack away from glass to the same extent as paraffin or pure shellac. This cracking away from glass, by the way, is probably an effect of difference in rate of expansion between the glass and cement which probably always exists, and, if the cement be not sufficiently viscous, must, beyond certain temperature limits, either produce cracks or cause separation. Professor Wright of Yale has used a hard mineral pitch as a cement in vacuum work with success. My attention has been directed to a fusible metal cement containing mercury, and made according to the following receipt, given by Mr. S. G. Rawson, Journal of the Society of Chemical Industry, vol. ix. (1890), P. 150:- Bismuth 40 per cent Lead 25 per cent Tin 10 per cent Cadmium 10 per cent Mercury 15 per cent This is practically one form of Rose's fusible metal with 15 per cent mercury added. It takes nearly an hour to set completely, and the apparatus must be clean and warm before it is applied. As the result of several trials by myself and friends, I am afraid I must dissent from the claim of the author that such a cement will make a really air-tight joint between glass tubes. Indeed, the appearance of the surface as viewed through the glass is not such as to give any confidence, no matter what care may have been exercised in performing all the operations and cleaning the glass; besides which the cement is rigid when cold, and the expansion difficulty comes in. On the other hand, if extreme air-tightness is not an object, the cement is strong and easily applied, and has many uses. I have an idea that if the joints were covered with a layer of soft wax, the result would be satisfactory in so far as air-tightness is concerned. This anticipation has since been verified.