sary is to take a weighed quantity of the substance, dissolve it, precipitate the whole of the silver by adding hydrochloric acid or other chloride till no more chloride of silver falls, collect the precipitate on a filter, wash, dry, and weigh. The amount of silver in the dried chloride, ascertained by calculation, is the amount of silver in the quantity of substance on which the operation was conducted; a rule-of-three sum gives the quantity per cent.—the form in which the results of quantitative analysis are usually stated. Occasionally a constituent of a substance admits of being isolated and weighed in the uncombined state. Thus the amount of mercury in a substance may be determined by separating and weighing the mercury in the metallic condition; if occurring as calomel (Hg,Cl) or corrosive sublimate (HgCl), the proportion of chlorine may then be ascertained by calculation (Hg=200; Cl=35.5).

So, then, a body may be isolated and weighed alone in a balance and its quantity thus ascertained; or it may be separated and weighed in combination with another body whose combining proportion is well known; this is quantitative analysis by the gravimetric method.

Quantitative analysis by the volumetric method consists in noting the volume of a liquid required to be added to the substance under examination before a given effect is produced. Thus, for instance, a solution of nitrate of silver of known strength may be used in experimentally ascertaining an unknown amount of a chloride in any substance. The silver solution is added to a solution of a definite quantity of the substance until flocks of chloride of silver cease to be precipitated: every 108 parts of silver added (or 170 of nitrate of silver: Ag=108, N= 14, 0,=48; total 170) indicate the presence of 35.5 of chlorine, or an equivalent quantity of any chloride. The preparation of standard solutions, such as that of the nitrate of silver, to which allusion is here made, requires considerable care; but when made, certain analyses can be executed with far more rapidity and ease than by gravimetric processes.

The quantitative analysis of solids and liquids often involves determinations of temperature and specific gravity. These processes will now be explained, after which an outline of volumetric and gravimetric quantitative analysis will be given. The scope of this work precludes any attempt to describe all the little mechanical details observed by quantitative analysts; essential operations, however, are so fully treated that expert manipulators will meet with little difficulty.

The analysis of gases and vapours also involves determinations of the varying pressure of the atmosphere, as indicated by the barometer (from Bapos, baros, weight, and μérpov, metron, measure), a glass tube 33 or 34 inches long, closed at one end, filled with mercury, and inverted in a cup of mercury. The mercury remains in the tube owing to the weight or pressure of the atmosphere on the exposed surface of the liquid, the average height of the column being nearly 30 inches. In the popular form of the instrument, the wheel-barometer, the cistern is formed by a recurvature of the tube; on the exposed surface of the mercury a float is placed, from which a thread passes over a pulley and moves an index whenever the column of mercury rises or falls. For further information concerning the influence of pressure on the volume of a gas or vapour, and for descriptions of the methods of analyzing gases, refer to Ganot's Physics" (translated by Atkinson), Miller's Chemical Physics,' and “Analysis of Gases" in Watts's 'Dictionary of Chemistry.'

6

MEASUREMENT OF TEMPERATURE.

As a rule, all bodies expand on the addition, and contract on the abstraction of heat, the alteration in volume being constant and regular for equal increments or decrements of temperature. The extent of this alteration in a given substance, expressed in parts or degrees, constitutes the usual method of intelligibly stating, with accuracy, precision, and minuteness, a particular condition of warmth or temperature-that is, of sensible heat. The substance commonly employed for this purpose is mercury, the chief advantages of which are that it will bear a high temperature without boiling, a low temperature without freezing, does not adhere to glass to a sufficient extent to "wet" the sides of any tube in which it may be enclosed, and from its good conducting-power for heat responds rapidly to changes of temperature. Platinum, earthenware, alcohol, and air, are also occasionally used for thermometric purposes.

The construction of an accurate thermometer is a matter of great difficulty; but the following are the leading steps in the operation. Select a piece of glass tubing having a fine capillary bore, and about a foot long; heat one extremity in the blowpipeflame until the orifice closes, and the glass is sufficiently soft to admit of a bulb being blown; heat the bulb to expel air, imme

diately plunging the open extremity of the tube into mercury; the bulb having cooled, and some mercury having entered and taken the place of expelled air, again heat the bulb and tube until the mercury boils and its vapour escapes at the open end of the tube; again plunge the extremity under mercury, which will probably now completely fill the bulb and tube. When cold the bulb is placed in melting ice. The top of the column of mercury in the capillary tube should then be within an inch or two of the bulb; if higher, some of the mercury must be expelled by heat; if lower, more metal must be introduced as before. The tube is now heated near the open end and a portion drawn out, until the diameter is reduced to about one-tenth. The bulb is next warmed until the mercurial column rises above the constricted part of the tube, which is then rapidly fused in the blowpipe-flame, and the extremity of the tube removed. The instrument is now ready for graduation. The bulb is placed in boiling water (a medium having, cæteris paribus, an invariable temperature), and, when the position of the top of the mercurial column is constant, a mark is made on the tube by a scratching diamond or a file. This operation is repeated with melting ice (a medium also having an invariable temperature). The space between these two marks is divided into a certain number of intervals termed degrees. Unfortunately this number is not uniform in all countries: in England it is 180, as proposed by Fahrenheit; in France 100, as proposed by Celsius (the Centigrade scale), a number generally adopted by scientific men: in some parts of the Continent the divisions are 80 for the same interval, as suggested by Reaumur. Whichever be the number selected, the markings should be continued above the boilingpoint and below the freezing-point as far as the length of the stem admits.

On the Centigrade and Reaumur scales the freezing-point of water is made zero, and the boiling-point 100 and 80 respec

tively; on the Fahrenheit scale the zero is placed 32 degrees below the congealing-point of water, the boiling-point of which becomes, consequently, 212. Even on the Fahrenheit system temperatures below the freezing-point of water are often spoken of as 66 degrees of frost ;" thus 19° as marked on the thermometer would be regarded as "13 degrees of frost." It is to be regretted that the freezing-point of water is not universally regarded as the zero-point, and the number of intervals between that and the boiling-point of water everywhere the same. The Centigrade scale, however, is being slowly adopted, and will doubtless, sooner or later, supersede the others.

The degrees of one scale are easily converted into those of another, if their relations be remembered, namely:-180 (F.), 100 (C.), 80 (R.), or 18, 10, and 8; or, best, 9, 5, and 4.

Formula for the conversion of degrees of one thermometric scale into those of another.

In ascertaining the temperature of a liquid, the bulb of a thermometer is simply inserted and the degree noted. In determining the boiling-point, the bulb is also inserted in the liquid, if a pure substance. In taking the boiling-point of a liquid which is being distilled from a mixture, the bulb of the thermometer should be near to, but not beneath the surface.

The following are the boiling-points of a few substances met with in pharmacy :

To determine melting-points of fats &c.-Heat a fragment of the substance (spermaceti or wax for example) till it liquefies, and then draw up a small portion into a thin glass tube, about the size of a knitting-needle. Immerse the tube in cold water

P