250 Proceediii(jfs of the Royal Irish Academy. 
the formation of a bluish white and basic precipitate. Therefore, for 
ilhistrating this experiment, it is better to use a slightly acid solu- 
tion, which, from the presence of the acid, is even more sensitive to 
heat, and regains its original condition on cooling ; hence this experi- 
ment can be performed with the same tube ad mfinitum. 
A solution of sulphate of copper, heated in sealed tubes, gives 
somewhat similar results as regards dehydration ; but a basic salt is 
determined even in an acidulated solution. 
The nickel salts do not, as a rule, so strikingly illustrate this phe- 
nomenon of chromatic change, although as it is well-known, the hydra- 
ted salts are green, and the anhydrous salts more or less yellow. 
Heated under pressure, the green chloride, when in solution, seems 
to be intensified in the first instance, then becomes yellow, and ulti- 
mately a dark yellowish green. This, at first sight, would appear 
somewhat anomalous, but, on examination, its anhydrous alcoholic 
solution presents similar anomalies on cautiously adding a weaker 
alcohol. 
It is necessary, in these experiments, that considerable caution should 
be used, as an explosion frequently occurs. As wire gauze would pre- 
vent the gradual change in colour from being observed, it is better to 
heat the experimental tube in a second glass, one considerably larger. 
The experimental, or hermetically sealed one, should be as small in 
calibre as the necessary observations will allow of. The volume of 
liquid necessary for proper observation will, of course, depend upon 
the depth of colour of the solution, and the amount of solution should 
have a small proportion to the capacity of the tube. Tubes for this 
purpose were made by taking a white glass tube of some considerable 
strength, and drawing it out until it has a diameter of y^-Q to y^^yths 
of an inch. My attention was first drawn to the value of capillary 
tubes for experiments similar to the above from a lecture delivered 
by Dr. Andrews, on the continuity of the liquid and gaseous state of 
matter." 
A most important observation made in connexion with this 
investigation is the difference observed between the efi'ects of dilution 
on colour- changes attending the basic condition, and on colour- 
changes attending dehydration. As might have been predicted on 
theoretical grounds, it is exactly the reverse in the latter case to 
what it is in the former. In a report recently read before the Academy, 
I pointed out that colour-changes resulting from formation of basic salts 
by dissociation (i. e. chromic or ferric salts) are influenced by the fact 
that dilution lowers the thermanalytic point. This is due to the basic 
action of the water itself. Eut it is self-evident that an increased 
volume of water will act differently in cases where the change of 
colour depends upon dehydration. In the first case the increase of 
volume in the water will assist the dissociation, and lower the ther- 
* Ni Clo 9, II2O, Marignac. 
