426 



Prof. J. J. Thomson and Mr. H. F. Newall. [Dec. 17, 



Notes. — (i.) Oleate of soda solution in water, made of that strength which best 

 gave certain results when a drop was let fall into a special sample of paraffin oil (see 

 below). 



(ii.) Sulphuric acid III, a mixture of acid and water, made by adding water to 

 strong acid until drops taken from the mixture and let fall into it gave rings as good 

 as many of the liquids in Class II. 



(hi.) Paraffin oil such as is ordinarily used in lamps. 



(iv.) Glycerine II, a mixture of glycerine and water, dilute enough to bring it 

 only just into Class II. 



(v.) Sugar solution in water, so strong as just to give rings when dropped into 

 itself. 



(vi.) Paraffin oil II, a special sample of oil that we have not been able to match, 

 and which will be referred to below (p. 432). 



(vii.) Sulphuric acid II, made of such strength that it should belong by the 

 character of its ring to Class III. 



(viii.) Sulphuric acid I, strong commercial acid. 



(ix.) Glycerine I, diluted with water, but still of such strength that rings were 

 never formed by drops of it falling into a column of it. 



These experiments indicate that the nature of the motion, after the 

 drop has fallen into the liquid, depends upon the value of fxjp. ISTow 

 fijp is a quantity of the dimensions of the product of a length and a 

 velocity, so that nothing can depend upon the absolute value of this 

 quantity, but only on the ratio of it to the product of some length 

 and velocity in the system. The most obvious length in the 

 system with which pjp can be connected is the size of the drop or 

 ring. If this is so, diminishing the size of the drop when fi/p is kept 

 constant ought to produce the same effect as increasing jxfp when the 

 size of the drop is kept constant. Now if we take a liquid from 

 Class I, the effect of increasing jx\p sufficiently would be to make it 

 behave like one in Class II, that is, give very much better rings, so 

 that for a liquid of this kind we should expect that small drops would 

 make better rings than large ones. To test this, we repeated the ex- 

 periments with the liquids of Class I, using much smaller drops than 

 we used before, and we found that now they made rings much more 

 readily and certainly, in fact, with this sized drop, they deserved to 

 be put into Class II. If we try the same thing with liquids in 

 Class III, we find that the rings are worse with small drops than 

 with large. On the other hand, increasing the size of the drop when 

 fijp is kept constant, should have the same effect as diminishing fxjp 

 when the size of the drop remains constant, so that if we repeat the 

 experiments with liquids from Class IV, using larger drops than 

 before, we should expect to get better rings. Experiment showed 

 this to be the case. Large drops were obtained by dipping a piece of 

 glass tubing into the liquid {e.g., strong sulphuric acid), and then 

 raising the tube after closing the top with the finger ; on removing 

 the finger, a considerable quantity of liquid flows through the tube, 

 making a large ring. The velocity with which /*//> would be most 



