122 Proceedings of the Royal Society of Edinburgh. [Sess. 
length of the illuminating portion of the lamp tube being about 11 cms. 
As may easily be seen, the lamp tube may be made in any convenient form, 
the essential parts of the lamp being the small vessel with its recess and 
the heating element situated in the recess. Thus a lamp suitable for 
illuminating purposes was constructed of the same kind of tubing as before, 
the tube being bent into the form of a spiral. This lamp was designed for 
250 volts pressure, the effective length of the tube being about 22 cms. 
It may be advisable to give here a short account of some of the more 
important points to be observed in the preparation of mercury vapour 
lamps. One of the chief precautions to be taken is to see that the lamp 
tube and the mercury to be put into it are perfectly clean and dry, for the 
presence of impurities or moisture interferes with the proper running of 
the lamp. A good method of cleaning the tube is as follows : — It is first of 
all thoroughly rinsed out two or three times with a fairly strong solution 
of chromic acid. Then it is washed several times with distilled water. 
Next, it is rinsed once or twice with absolute alcohol, and finally with 
ether. It is then fixed to a pump and exhausted, being slightly warmed 
after a time, in order to drive off the last traces of ether vapour. 
The mercury, which should be of the purest quality, may be introduced 
into the lamp through a side tube which serves to connect the lamp to the 
pump. A very good plan to adopt is to constrict this side tube at two 
places, one next the lamp and the other a short distance from it. The 
object of this will be seen immediately. The lamp, after being connected 
to the pump, should be evacuated to a fairly high degree of exhaustion,* 
and should be strongly heated during the process, in order to facilitate the 
removal of any volatile gases that may be present. One or more tubes of 
concentrated sulphuric acid, calcium chloride, or phosphorus pentoxide 
should be introduced between the lamp and the pump, in order to absorb 
any moisture that may be given off. When the lamp is sufficiently 
exhausted, it should be sealed off at the constriction of the side tube nearer 
the pump. The lamp should then be connected to a suitable battery and 
run for some time, after which any gases that may still have been given off 
may be worked up into the side tube, which should then be sealed off at 
the other constriction. 
The experimental lamps made of Jena glass naturally proved to be more 
satisfactory than the previous ones made of soda glass, because, not only 
could the method of starting the arc be observed, but also the lamps could 
be run for considerable periods, because Jena glass can withstand the heat 
of the lamp when burning. As in the other case, the arc was struck about 
* A Sprengel three-fall pump was used by the authors. 
