170 Scientific Proceedings, Royal Dublin Society. 



could now be made with about the same percentage accuracy as obtained for 

 the larger intervals given in Table I. 



A great part of this work was completed more than a year ago, but the 

 hope that it would be possible to push the investigation to still shorter 

 intervals caused publication to be held back. It appears, from our own 

 experience, and in view of the investigations of Klopsteg and of Webster 

 already mentioned, that the " capacity " methods of measurement for short 

 time intervals can be relied on under favourable conditions to measure to, at 

 least, the nearest micro-second. In our first arrangement larger errors than 

 this occurred, but the improved mounting seemed to promise readings having 

 this accuracy. Under extremely favourable conditions, and using some 

 special apparatus, Klopsteg reduced the errors to one-tenth of this. 



