Carbonic Acid on Smooth Glass Surfaces, 1 65 



and the threads contained in these were spread on the ordinate 

 plate of the microscope-stage and counted. These, with the ends 

 of the skeins cut off, gave for 2488 threads of 2 metres length 

 a weight of 2*6529 grms. The following comparison of the 

 number deduced in this way and that arrived at independently 

 from the diameters of a similar number of threads, shows 

 what degree of accuracy can be attained by the method 

 adopted. The weight of 1877*4 threads, of 1000 millim. long, 



by direct weighing 1*0025 grm. 



by estimation of diameter... 1*0000 „ 



In previous investigations on the condensation of gases on 

 the surface of glass, as well as in researches with other solid 

 bodies, it has been taken for granted that this condensation 

 became quite stationary in a few hours or days. If then this 

 supposition is erroneous, the value of all the results deduced 

 from such investigations will be called into question; and so it 

 appeared to me necessary to determine the dependence of the 

 condensation of gas on time more correctly than had yet been 

 done. For this purpose I have made use of the apparatus repre- 

 sented in PI. V. fig. 1. This consists of a wide measuring-tube, 

 provided with an etched millimetre-scale throughout its whole 

 length, closing air-tight with a stopcock at its upper end, and 

 communicating with a wider vessel, A, for the reception of the 

 glass threads. In the upper part of this vessel is the stopper C, 

 ground in air-tight, and communicating with the outer air by 

 the glass stopcock, c. The whole of the joints, made tight with 

 a very consistent mixture of wax and grease, were proved to be 

 perfectly impervious by observations extending over three 

 years. The measuring-tube B was carefully calibrated and 

 the error of meniscus estimated ; the capacity V of the empty 

 vessel A, from the closed stopcock b to the stopcock c, was 

 measured with mercury, as well as that of the passages in the 

 stopcocks. The volume V is diminished in the experiments 

 by that of the enclosed glass-threads. Take sg as the specific 

 gravity of the threads contained in the vessel A, and p as their 

 weight, then the space remaining unoccupied by them, viz. 



V„=V-^ (4) 



In order to be able to alter the pressure in the apparatus 

 during the experiment, the mercury-trough had fitted into it 

 a tube of such a depth as to allow- the measuring-tube to be 

 sunk in for its whole length. The filling of the apparatus 

 and the observations were made in the following way : — The 



