TRANSACTIONS OF SECTION B. 549 



On the so-called ' Normal ' Solutions of Volumetric Analysis. 

 By Alfred H. Allen, F.C.S. 



3. On the Determination of the Loss of Heat in Steam-Boilers arising 

 from Incrustation. By William Thomson, F.B.S.E. 



On evaporating large quantities of water in small vessels I was surprised to find 

 that some samples evaporated away mucli more rapidly than others, although the 

 same sizes of flames were allowed to play on the bottoms of each of the vessels ; 

 and I afterwards found that the rapidity of evaporation was in the inverse pro- 

 portion to the quantity of incrustation which had formed on the bottoms; I 

 was further impressed by the small quantity of incrustation which was required 

 to make a great difference in the quantity of water evaporated. Based on this 

 fact, I have constructed an apparatus in which four small vessels are set so that 

 the bottom of each dips into paraffin contained in a large vessel, heated by 

 one lamp from the centre ; by this means the same quantity of heat is given to 

 each vessel, and it is evident that on measuring the quantity of water evaporated 

 by each in a given time, the amount of heat lost by the incrustation may be 

 determined. 



In the arrangement of this apparatus I use one of the small vessels to 

 evaporate pure distilled water, and the others to evaporate waters to be tested for 

 boiler purposes ; these samples are placed in flasks or vessels arranged mouth down- 

 wards, on Bischofl"'s principle, so that the water in the small vessels is always 

 kept at the same height until the whole in the flask is evaporated. When the 

 distilled water is evaporated to a certain point the operation is stopped, and the 

 amount of water left is in each case accurately measured. It is evident that the 

 quantities of the different samples which remain in excess of the distilled water 

 would represent in water the quantity of heat lost by the incrustation produced by 

 a given quantity of any sample. 



4. On the Identification of the Inh used in ivriting Letters and Documents 

 as Evidence in Cases of Libel, Forgery, ^c. By William Thomson, 

 F.E.S.K 



Most of a large number of inks on envelopes sent to the author by different 

 persons, when tested by reagents, were found to differ from each other to a large 

 extent, whilst no two were precisely similar when minute compaiisons were made. 



The reagents which I have found to give best results are : — 



1. Dilute sulphuric acid. 



2. Strong hydrochloric acid. 



3. Slightly diluted nitric acid. 



4. Sulphurous acid solution. 



5. Caustic soda solution. 



6. Cold saturated solution of oxalic acid. 



7. Solution of bleaching powder. 



8. Solution of protochloride of tin. 



9. Solution of perchloride of tin. 



As an example of the results produced by reagents, some inks when treated, for 

 instance, with sulphuric acid are changed to bright crimson, some to deep red, 

 whilst others become blue, green, violet, or grey, of different shades, and some 

 remain practically unaltered ; and when, as sometimes happens, the same or nearly 

 the same colour is produced by one reagent, the colours produced by others are 

 very different. 



By testing inks in this way on libellous letters, or in cases of forgery, the ink 

 may, under some combinations of circumstances, possess as much individuality 

 about them as the faces of their owners. 



I have found that the same ink when put on different kinds of paper gives pre- 



