14 Journal of the Asiatic Society of Bengal. [N.S., XVI, 
the effect of the sulphuric acid could only be relied on to keep 
the leak in wet weather below the standard value for some- 
thing like 200 to 300 hours. It was also found that the effect 
of the moisture absorbed by the wooden framework of the case 
could not be neglected. With proper hermetical sealing and 
better arrangements for introducing drying materials into the 
case, it was thought possible that better results could be ob- 
in y substituting for strong sulphuric acid other desic- 
cants such as Phosphorus Pentoxide or metallic sodium 
these problems have since been investigated. A square 
brass case, free from wood, was constructed, and after much 
difficulty rendered absolutely air-tight by soldering all edges 
and stopping the minute leaks with coatings of Canada Balsam 
in ether. The threads operating the key were led in through 
small U tubes containing mercury, the communication being 
rendered perfectly air-tight by the addition of oil. Since the 
oil increased friction when cotton threads were used, these 
were replaced by human hair, as experiment showed that this 
was the material which worked most smoothly. The sulphuric 
acid was introduced or removed by a double siphon without 
disturbing the sealing of the case. To eliminate all possibility 
of the entry of moist air by leakage a slight additional pres- 
sure was maintained inside the case. Drying agents other 
than sulphuric acid were introduced from the top of the case, 
Science Congress at Lucknow, but have not hitherto been 
published. 
As a result of the experience obtained during these investi- 
v 
recording registration of the Earth-air current, have been in 
he outer case is now made cylindrica! instead of square, 
thus avoiding many difficulties in construction, especially by 
reducing the number of joints, all of which are possible sources 
