ON A QUARTZ THREAD GRAVITY BALANCE. 225 



behaved so badly that we actually took the trouble to assure ourselves by trial of ita 

 sujieriority over other materials. We mounted a very fine steel wire, kindly sent to 

 us by Mr. EL.LERY of the Melbourne Observatory, on another balance, arid observed 

 that its rate of viscous subsidence was about a hundred times greater than we were 

 accustomed to in the case of quartz. We repeated the experiment with a very fine 

 platinum wire with a similar result. 



It must not be supposed that all fused quartz, as derived from clear rock crystal, 

 has the same properties. Almost every crystal examined by us contains both sodium 

 and lithium the latter in large spectroscopic quantity indeed, we first noticed it 

 from the colour it gave to the blowpipe flame. The observation of the almost universal 

 presence of lithium in quartz was first made by TEGETMEIER,* a fact of which we 

 were ignorant when we made the observation. There also appeal's to us to be a distinct 

 though small difference in the viscosity of various samples of fused quartz, and this 

 independently of the sodium or lithium they may contain. It has been our practice 

 to select the most infusible quartz independently of the amount of lithium it contains, 

 for lithium seems to be burned out by continued heating in the oxy-gas flame : 

 whether this is really the case or whether the lithium forms a compound which does 

 not give the flame re-action, we have not attempted to inquire. We have aimed at 

 securing the greatest possible uniformity in the thread and a mean diameter of about 

 0038 centim. We do not pass a thread unless the diameter is uniform from end to end 

 within the limits of observation (exceeding those of measurement), looking at samples 

 taken from each end of the thread through a microscope magnifying 100 diameters. 

 We have also spent a great deal of time in trying to make certain that there were no 

 drawn-out air bubbles in the thread, and though we have succeeded in arriving at 

 the satisfaction of both these conditions simultaneously, still our present practice is 

 to ignore very small bubbles, and to direct our attention principally to obtaining 

 uniformity. Since, however, we now adjust the torsion of one end of the thread 

 only, we are inclined to think that for the future we would put up with a slight taper 

 in the thread, provided of course that the twisting for adjustment was done from the 

 thinner end. 



The bow-and-arrow method of Mr. BOYS gives better threads for our purpose than 

 the catapult method (' Laboratory Arts '), which does not lend itself to the production 

 of threads of great uniformity. There may also be some difference of tempering, 

 owing to the different rates at which the threads are pulled out and cooled in the 

 two methods considered. A great deal of time may be spent in preparing the thread, 

 for at present we have no method of predetermining its diameter. All that we can 

 do is to be guided by experience, and shoot threads till one is got satisfying the 

 conditions. The process is so uncertain that we have on occasion got a thread within 

 a few days, and on others we have spent a fortnight over it. The thread which forms 

 part of the instrument to which the observations refer took a fortnight's continuous 



' Wied. Ann.,' 41, p. 19, 1890. 

 VOL. cxcin. A. 2 a 



