The cause of greasiness in burettes, a perpetual source of trouble in accurate work, has been discussed by various writers, and attributed by them to oils or fats in the distilled water derived from the lubricating materials of the still (Ostwald), or to the vaseline, &c., used for lubri- cating the stop-cock. Though probably both causes may be at work, yet we consider that this greasiness is produced also by slime-forming bacteria or hypomycetes, especially Dematium, the spores of which settle and develop on the sides of the vessel ; our experience has shown that if the air sucked in by the outflowing solution be filtered through cotton-wool plugs, and if the burette be filled invariably from below, and not, as is customary, from above, its state of cleanliness may be greatly prolonged.
The first table of specific gravities of nitric acid in terms of per- centages was constructed by Kirwan ;* subsequent tables were published by Ure,f Thenard and Kolb.J The numbers given by the last are still reproduced more or less fully in text-books of chemistry and physics, though it is now generally admitted that his results for the more concentrated acids are too low.
But as a considerable advance over previous writers, Kolb pointed out, firstly, that the presence of the lower oxides of nitrogen alters appreciably the value of the density ; secondly, that acid of approxi- mately 100 per cent, concentration has a high coefficient of expansion; and thirdly, that the maximum contraction on admixture with water occurs at the percentage composition required for the hydrate HN0 3 .3H 2 0. Upon the last result Kolb expressed himself as follows : " Ce fait d'un maximum de contraction correspondant exactement a un melange en proportions definies n'est pas un effect du hasard." The most recent work upon the densities is that published by Lunge and Key, which is now more generally quoted in technical manuals.
The analytical methods of these last writers are similar in outline to those herein adopted, but their final process of purification consisted in distilling the concentrated acid with sulphuric acid in vacua, a method which might cause slight contamination, as the vapour pressure of the latter acid under these conditions could not be an insignificant quantity.
The densities were determined for 15/4, but it is not evident as to those actually made, and those interpolated in order to reduce the values to the commercial scale of Twaddell ; further, the results are only taken to 0*001, which is far within the limits of accuracy attainable.
These writers, in conjunction with Marchlewski, have given a table
- 'Irish Acad. Trans.,' vol. 4, p. 3 (1790).
t ' Quart. Jourii. Sci.,' rol. 4, p. 291 (1818).
t 'Ann. Chilli. Phjs.,' [4], vol. 10, p. 140.
' Zeits. f. Angew. Chemie,' 1891, p. 167, and 1892, p. 10.
