494 Messrs. T. and A. Gray and J. J. Dobbie. [Apr. 24, 



seems high when the quantity of lead oxide is considered, while the 

 density of XX seems low for the same reason. Also comparing X 

 with XX, we see that while XX contains more lead than X, and at the 

 same time less of impurity, XX has a considerably lower density and 

 a much lower specific resistance. The determinations of density and 

 resistance were repeated for XX, and the results found practically to 

 agree with those previously obtained. It may be interesting to note 

 that, while all the others were new glasses, X and XX were glasses at 

 least ten years old, which had been made for Sir William Thomson's 

 experiments on the preservation of electric charges for a long time 

 in hermetically sealed glass vessels. 



The flask marked VII has a somewhat lower percentage of lead 

 oxide than any one of the other glasses which are similar in resisting 

 quality, but it will be noticed that its density is the highest of all. 

 Only a determination of density and of the percentage of lead have 

 been made for this glass. In the case of No. IV, the resistance was 

 determined, but no analysis was made. We ought to state that 

 I, II, III, IV, were all made by Messrs. Osier, Birmingham, and 

 were understood to be from the same pot of glass. 



Referring to the table on page 493, and considering the relation of 

 the observed resistances to the amount of soda contained in the glass, 

 we see that, except II and the anomalous glass XX, the resistances 

 decrease as the percentage of soda increases ; and making the same 

 comparison for the percentages of potash, we find approximately the 

 same kind of variation of resistance. This seems to indicate that, if 

 other things were equal, the resistances would be diminished by in- 

 creasing the proportion of alkali, a result which agrees with previous 

 observations. We do not, however, consider the effect of alkali to be 

 so important as the opposite effect of lead oxide. 



In the paper referred to above (" Proc. Roy. Soc," vol. 34, p. 199) 

 it is stated that a glass which seems to approach in composition to a 

 definite chemical compound is also good in point of resistance. It seems 

 possible that in the case of those glasses which have approximately 

 the same composition but different densities, those having the higher 

 densities may approach more nearly to definiteness of composition. 

 At all events, it is well known that minerals, more or less resembling 

 glass in composition, which crystallise, have a greater density than 

 substances of similar composition which are non- crystalline. 



It will be seen that the glasses which we have experimented on 

 differ very widely from those analysed by Dr. Divers. Although it 

 is impossible to say in what proportion the silica is distributed 

 among the bases, it may be useful to give the number of combining 

 proportions of the oxides present in the various glasses, and the 

 simplest formulas which can be deduced from the analyses. 



