certain Dilute Aqueous Solutions. 395 



crosses in circles respectively. It will be noticed that the 

 two series for potassium carbonate do not coincide but lie 

 parallel to one another. If there were any impurity in the 

 body of the water that decreased its surface-tension, it 

 could not be removed by the flexible hoop, and probably it 

 would lower the whole curve as we observe in this case. 

 Those points for zinc sulphate which are represented by 

 crosses in circles were obtained from a series of readings on 

 solutions of the anhydrous salt. This contained quite a little 

 of the sulphide, which floated around in the solution but 

 apparently produced no effect on the wave-length of the 

 ripples. The values obtained by Sentis are represented by 

 circles. 



It is evident that these results can be best represented by 

 straight lines. If we write Tg = Tw + kc, where Tg is the surface- 

 tension of the solution containing c gram-equivalents per 

 litre, and Tw is the surface-tension of water at the same 

 temperature, and tc is a constant, we can find k from the 

 plotted results. The values found for k by various observers 

 are given in the following table ; those assigned to Sentis 

 were calculated from his observations on solutions of 0*56 

 and 1*13 normal concentrations. 





Yolkmann. 



Quincke. 



Sentis. 



Rother. 



Dorse 



NaCl ; k= 



1-59 



1-57 



1-57 



1-38 



1-53 



KC1; k = 



141 



1-57 



1-41 



147 



1-71 



i(Na 2 C0 3 ); ie= 



0-987 



1-57 







2-00 



KK 2 C0 3 ); *= 



1-78 



1-57 







1-77 



l(ZnS0 4 ): k= 







1-78 





1-86 



All of Yolkmann's curves appear to become steeper as they 

 approach zero concentration • and his curves for K 2 C0 3 and 

 Na 2 C0 3 cross near the origin, so that the curve for Na 2 00 3 

 for dilute solutions becomes steeper than the one for K 2 C0 3 . 



The comparison of the results of this work with those ob- 

 tained by other experimenters must not be pushed too far, 

 because they have worked exclusively on solutions of greater 

 concentration than half-normal, while most of my observa- 

 tions are on solutions less concentrated than that. 



The average of all observations taken on water gives 

 T 17 o. 44 =73*25, or Ti 8 °= 73*17, which is rather low. If we 

 omit the values Ti 8 ° = 72*77, 72*84, and 72*76, which are 

 certainly too low, we find the averages given in the last line 

 of the table ; viz., T 17 o. 48 = 73*32, which gives T 1S o = 73*24 and 

 T = 75*98. This is lower than the values found by Quincke, 



