Glass Bulbs in Liquids. 29 



The last two numbers suggest the same remark as in (4). 



6. A sample of hydrocarbon-oil was next tried, a different 

 weight and bulb being used. Each result is a mean of six 

 determinations. The sp. gr. of the oil was '9120 at 16°, i. e. 

 the same as that of the mixture of alcohol and water referred 

 to in (4). 



Table III. 



d. t olbs. t calc. 



5-66 31-5 31-41 



4-76 35-3 34-91 



3-37 46-0 46-80 



2-79 57-4 57-73 



[2-20 90-8 78-88] 



Equation (^-22-96)^=270-65. 



Probable error of a single (1-4) experiment, 0*37. 



II. Change of the Unbalanced Pressure. 



7. The unbalanced pressure, to which the ascent of the bulb 

 is due, can be conveniently changed by increasing or dimi- 

 nishing the weight attached to the bulb. If we take a centi- 

 metre as the unit of volume, and understand by 8 the mean 

 density of (bulb -f contents + weight) referred to that of the 

 liquid as unity, then (1 — S)=p may be termed the " unba- 

 lanced unit pressure." 



The following experiments were performed with a bulb of 

 the usual size, recently-boiled water at 11 0, 5 being the liquid. 

 The diameter of the cylinder was 5*66 centim. ; the times are 

 stated in half-seconds; and each result is the mean of seven 

 determinations. 



Table IV. 



t. p. tp*. 



20 -01139 -0025946 



40 -00810 -0026244 



81 -00569 -0026225 



Equation, tp 2 = c= '0026138. 

 Probable error of c, 0'25 per cent. 



8. Another set of experiments was carried out in the same 

 cylinder, under nearly the same conditions, with rive varia- 

 tions of weight. Each result is the mean of five observations 

 of ascent through a height of 16*5 centim. The times are 

 stated in thirds of seconds. 



