A Roman Glass Bottle 239 



1890. March I3th, 385th meeting. Mr. Evans described 

 a small glass bottle (about three inches in diameter) of 

 Roman make, inside which five glass rods extended from the 

 base of the neck to a point a little above the bottom, at 

 angles of about 45 : with the axis of the bottle, and thus 

 seeming to support the neck. He suggested that, while the 

 bottle was still soft and before it had reached its full size,, 

 a blunt rod, like a knitting pin, had been pushed into the 

 side at live different points and the viscous glass carried as 

 tubes to the base of the neck, where they were finally attached 

 by the help of a blowpipe. This done, the bottle would 

 be reheated and blown to its full size, the tubes by this 

 action being welded into rods. By adopting this device, 

 Mr. Powell, whose practical knowledge of glass is so great, 

 and Mr. C. V. Boys, so successful in constructing glass 

 apparatus, had made a similar vessel. 



May Sth, 3S/th meeting. Professor Judd exhibited a 

 crystal of beryl, well formed, though somewhat water-worn, 

 weighing 2650 grains, with a specific gravity 2-703. It was 

 clear, and its colour between aquamarine and emerald. 

 It had been obtained in Ceylon by Mr. Barrington Brown. 

 He had also brought from Burma the specimens (exhibited) 

 of rubies in a matrix of highly crystalline limestone. 



Oct. 30th, 38gth meeting. Professor Dewar exhibited 

 a specimen of liquid nickel carbon monoxide, discovered 

 by Mr. Mond while experimenting with an improved form 

 of Grove's gas-battery in order to find a way of obtaining 

 hydrogen cheaply for commercial purposes. While trying 

 various modes of separating the two oxides of carbon from 

 their mixture with hydrogen in the gas-battery he found 

 that nickel united with the carbon monoxide in the pro- 

 portion of one volume of the former to four volumes of the 

 latter, forming a gas at ordinary temperatures, which is 

 reduced by pressure to a heavy, colourless, highly refracting 

 liquid. As neither cobalt, nor iron, nor any other metal 

 forms a similar compound, this gives a ready method of 

 distinguishing nickel from any metal resembling it. That is 

 the more important, because nickel and cobalt have the 



