126 PROCEEDINGS OF THE AMERICAN ACADEMY. 



undertaken. A thoroughly satisfactory window is obtained in this 

 way.i* 



After the glass vessel has been hermetically sealed, the soft iron 

 armature, J, can be turned by means of a magnet fixed suitably out- 

 side. The magnet is placed on a circular platform of brass surrounding 

 the tube so that the poles of the magnet control the armature within. 

 The platform is supported on the top of the box containing the glass 

 part of the apparatus. The platform has a circular groove in it which 

 a circular brass ring fastened to the lower side of the magnet exactly 

 fits, so that the magnet may be slid around without altering its position 

 with respect to the tube. 



To put this apparatus together required some care in handling, and 

 no little skill in glass-blowing.^^ The large cylinder of glass was blown 

 about 40 cm. long, and when the three supporting wires, one of which 

 is shown at 0, had been sealed in, the lower end of the cylinder was 

 opened. To the upper end the narrower tube was joined, and the short 

 tube bearing the window, D, was then sealed on at the i)roper place. 

 The next step was to fasten the fibre so that it might be heated to a 

 temperature of, say, 300° C, without danger of slipping or breaking. 

 Various attempts were made. For example, the end of the fibre was 

 platinized, and electroplated with copper to the supporting wire, — a 

 very troublesome operation. It was found, moreover, that when heat 

 was applied a break occurred where the quartz came in contact with 

 the metal, so that this method was abandoned. Carbon cement was 

 also tried, but discarded owing to the uncertainty whether all of the 

 volatile substances which it contains were driven off. Clamping was 

 resorted to, and the simple form of clamp shown in the figure adopted. 

 With the clamp, the danger of snapping the fibre just at the edge of 

 the metal is considerable. To minimize this, platinum foil was wrapped 

 around the part of the fibre to be placed in the clamp. The upper 

 large plate was now pushed on to the wire, B, and the small disk 

 clamped to the wire in the manner already described. The lower 

 plate was then placed on a support, the glass separating pieces placed 

 temporarily upon it, and the disk suspended in position between 

 the plates. A short range horizontal telescope was then focussed on 

 the edge of the suspended disk, so that, when the latter was made to 

 revolve, one could tell very readily if the wire, B, had been fixed at 

 right angles to the disk. In order to remedy any want of perpendicu- 



" This work was done by Mr. Lundin of the Alvan Clark Optical Company, 

 Cambridge. 



^^ Mr. Gelling, the glass worker of the Jefferson Physical Laboratory, has 

 done the glass-blowing with great care and efficiency. 



