103 
sealed with safety. It is advisable to number the bulbs, and 
to test the first and last by exposing them to a strong light. 
Frequently, in spite of every precaution, the gas explodes 
during the act of sealing, so that in this operation it is advi- 
sable to hold the bulb with a cloth rather than in the open 
hand. As soon as one bulb tube is removed another is 
placed in connection with the evolution flask, and after ten 
minutes sealed as described. The above quantity of acid 
will serve for the production of sixty bulbs. Thus prepared 
the sealed bulbs may be bept in the dark for any length of 
time without injury; some, which were known to have been 
made more than a year, were found to be perfectly good. To 
explode these bulbs it is only necessary to expose them to 
diffuse daylight or sunlight, when the combination occurs 
instantly. Of artificial lights the bright flash produced by 
the combustion of the vapour of bisulphide of carbon in nitric 
oxide is most effective ; but the light of burning magnesium 
wire, of phosphorous in oxygen, or the electric light, answers 
perfectly well Professor Roscoe stated that Mr. Dancer, of 
Cross Street, had undertaken to supply the bulbs to persons 
unable to prepare them. 
A paper “ On a new form of roof for Dyehouses,” by Mr. 
John Thom, was read, communicated by the President. 
The object of the present communication is to describe the 
construction of a roof for buildings in which there is a good 
deal of vapour, so as to produce the minimum amount of con- 
densation of such vapour inside the buildings, and thus avoid 
the production of drops, as well as the minor evil of an 
obscure atmosphere. Any one practically acquainted with 
dyeing processes knows well the loss which arises from these 
two evils. In all cases that I have seen, except where the 
dyehouse was the lower fiat of a series, and where, of course, 
the covering of the dyehouse was the floor of the room above, 
the roof of a dyehouse is made about the usual angle or 
