Amongst the phenomena which characterise different bodies before the blowpipe, and serve for their distinction, the colour of the flame is of no small importance. This power of colouring the blowpipe flame is not, however, exhibited by all bodies with sufficient intensity to enable them to be distinguished by it with certainty; and certain substances are consequently usually employed, such as muriatic acid with baryta, strontia and lime, or sulphuric acid, partly to form and partly to set free volatile compounds. By this means, however, although the intensity of the coloration is heightened, its duration is not increased, as these acids, and particularly muriatic acid, evaporate for the most part before they have acted sufficiently, so that the coloration lasts only for a few moments. This defect may be got over by the employment, instead of the volatile muriatic acid, of a chloride, which will retain the chlorine at a high temperature, so that it may only be set free by degrees in small quantities, while the body forming its base may be without action upon the colouring power of the body under investigation. For this purpose chloride of silver appears to be the best, especially as it may readily be prepared in a state of purity. The best plan is to stir it with water into a thick paste, and keep it in a bottle.

In regard to the action of chloride of silver upon the coloration of the blowpipe flame, I have investigated. several compounds of potash, soda, lithia, lime, baryta, strontia, copper, molydenum, arsenic, antimony, and lead, and mixtures of these substances. Chloride of silver, of course, has no action upon borates and phosphates, both of these acids being amongst those which offer the most resistance to the action of heat.

For a support, I employed first of all platinum wire, but this is soon alloyed by the metallic silver which separates, and thus rendered useless in testing metals. Silver wire is too readily fusible, and also difficult to obtain free from copper, which may give rise to errors when in contact with chloride of silver. For these reasons, iron wire is best fitted for experiments with chloride of silver, as from its cheapness a new piece may be employed for each experi

ment, while the silver may readily be obtained in the form of chloride from the broken pieces. If the size of the fragment under examination be sufficient, the platinum forceps may be employed.

The results at which I arrived, by the employment of chloride of silver, in comparison with those obtained without this reagent, are as follows:

'With potash compounds, such as saltpetre, potashes, &c., the flame is decidedly of a darker colour with chloride of silver; and even in ferrocyanide of potassium, which, when treated by itself with the blowpipe, colours the flame blue, the addition of chloride of silver produces a distinct potash coloration.

The action of chloride of silver upon soda salts is not so favourable; for although with some, as nitrate of soda, common soda, and labradorite, the flame acquires a more intense yellow colour by the addition of chloride of silver, this reagent produces no observable difference with other soda compounds, such as sulphate of soda and analcime. This also applies to the compounds of lithia, some of which give a finer purple-red colour on the addition of chloride of silver, whilst upon others it has no such effect.

6

With lime compounds chloride of silver acts favourably upon the colouring power. Thus the addition of chloride of silver to calcareous spar or gypsum (in the reduction flame) gives the flame a more distinct yellowish-red colour, but stilbite gives no coloration either with or without chloride of silver. With fluor-spar the coloration cannot well be observed, as this decrepitates too violently under the blowpipe.

6

The action of chloride of silver upon compounds of baryta and strontia is decidedly advantageous, as both the intensity of the coloration and its duration leave nothing to be desired. Siliceous celestine, which, when heated by itself in the forceps, scarcely coloured the flame, immediately produced a permanent red coloration when heated with chloride of silver.

Although it appears from the preceding statements, that the employment of chloride of silver presents no advantage

with some substances, it may be used with good results in the treatment of mixtures of alkalies and earths.

Thus, with petalite alone, the lithia coloration is first produced, and a slight soda coloration is afterwards obtained; whilst with chloride of silver the soda coloration appears very distinctly after that of the lithia. With lithion mica alone a very distinct lithia coloration is presented; but in the presence of chloride of silver a colour is first produced which may lead to the conclusion that potash is present, but the lithia coloration is weakened. Ryacolite, heated by itself in the blowpipe flame, only gives a distinct soda coloration; but with chloride of silver a slight potash coloration is first produced, and the colour of soda then appears very distinctly; the lime contained in it cannot however be detected by the coloration of the flame.

Chloride of silver may be employed with still greater advantage with the following metals, but in these cases it is particularly necessary that the operator should become familiar with the colour produced by each individual substance.

'With copper compounds, such as red copper ore, malachite, copper pyrites, sulphate of copper &c., when contained in other minerals so as to be unrecognisable by the eye, the employment of chloride of silver may be of the greatest service, as the smallest quantities of copper, when treated with chloride of silver under the blowpipe, give a continuous and beautiful blue colour to the flame. With chloride of silver the presence of copper may be distinctly ascertained by the blowpipe, even in a solution which is no longer coloured blue by the addition of ammonia.

"The employment of chloride of silver will be equally advantageous with molybdenum, as in this case also the flame gains greatly in intensity. Arsenic, lead, and antimony are already sufficiently characterised, the former by its odour, the two latter by their fumes; but even with these metals chloride of silver may be employed with advantage to render their reactions still more distinct. It is only necessary to observe, that the greenish-blue fame of antimony appears greener and more like that of molybdenum under the influence of chloride of silver.

'Chloride of silver may also be employed with compounds containing several of the above-mentioned metals.

'If bournonite be heated in the oxidation flame of the blowpipe, a fine blue flame is first produced, which indicates lead with certainty; if chloride of silver be now applied, copper is also readily shown. The antimony contained in bournonite cannot be ascertained by the coloration of the flame; but this may easily be detected upon charcoal, or in a glass tube open at both ends.

'Native molybdate of lead, without chloride of silver, only gives a blue colour to the blowpipe flame; with chloride of silver this blue coloration of the lead comes out more distinctly, but at the same time the tip of the flame, particularly when the reduction flame is employed, appears of a beautiful yellowish-green colour from molybdenum.

'With mixtures of arsenic and copper, or antimony and copper, the flame first acquires a greyish-blue or greenishblue colour from the oxidation of the arsenic or antimony; the copper may then be very easily detected by chloride of silver. This applies also to mixtures of arsenic and molybdenum, or antimony and molybdenum; with chloride of silver the yellowish-green flame of molybdenum appears distinctly. It will be more difficult to analyse mixtures of arsenic and lead, or antimony and lead, in this manner; and if a compound contain both arsenic and antimony, these two bodies are not to be distinguished with chloride of silver under the blowpipe.

From these experiments it appears that in blowpipe testing, it is more advantageous to employ chloride of silver, instead of muriatic acid.

'Chloride of silver is particularly to be recommended in testing metallic alloys for copper. Thus, to test silver for copper, chloride of silver may be applied to the ends of silver wires, and on the application of heat the smallest quantity of copper will furnish the most distinct reaction. This is as sensitive as any of the known copper reactions, and may be performed quickly and easily. In testing metallic alloys for traces of copper, it may be advisable to submit those which contain antimony, zinc, lead, and other volatile

metals to roasting, so as to drive off these metals before the addition of chloride of silver.'

GENERAL ROUTINE OF BLOWPIPE OPERATIONS.

Size of the Assay.-The morsel operated on is sufficiently large when the effect of the heat and the fluxes added can be distinctly discerned. The size of the assay-piece generally recommended is much too large; its size ought to be about that of a mustard-seed; that of the flux added, about the size of a hemp-seed. It should in general be previously reduced to fine powder.

6

SODA-PAPER.-Mr. Forbes writes as follows in the Chemical News: As it would be impossible to submit any powdered substance to the direct action of the blowpipe flame without its suffering mechanical loss, some means must be employed for holding the particles together until they are so agglutinated by the heat that no such loss need be apprehended; this is secured by the use of the soda-paper envelope or cornet, as devised by Harkort. For this purpose slips of thin slightly sized writing paper, about 14 inch long

g

Fig. 4.

FIG. 70.

by 1 inch broad, are steeped in a solution of one part crystallised pure carbonate of soda (free from sulphate) in two parts of water. When dried these are used for forming small cylindrical cornets, by rolling them round the ivory cylinder, fig. 70 d, previously described. A hollow is formed to them by folding down a portion of their length on to the end of the cylinder, which is then pressed firmly into the corresponding mould in the blowpipe anvil, and which, upon the withdrawal of the cylinder, serves as a support until they are filled with the assay from the scoop in which the assay and flux have been mixed. After pressing the assay down,

a{