such a nitric acid, it should be placed in a well-stoppered bottle while. weighing.

Second, such suggestions as not placing the substance to be weighed directly on the metallic pans, and keeping the pans free from all moisture, seem unnecessary to one who appreciates the value of a balance, and convenient for accomplishing these results are the movable glass scale pans which can be purchased from any scale manufacturer.

Third, weights should never be left on the balance, and, whenever possible, should be transferred to the balance by means of pincers.

Fourth, under no circumstances should the balance be left to oscillate after weighing is finished, and during the weighing it should be brought to a state of rest whenever new FIG. 15.—Model of torsion balance. weights or a new substance is being placed

thereon.

WEIGHTS

The weights used on scales and balances are composed of iron, brass, or aluminum, and of such shape as convenience suggests. Thus, the cheap avoirdupois weights are made of iron, while those finer are composed of brass, and the more expensive are nickel plated, and their shapes are usually those shown in Fig. 17.

Fig. 16.-Effect of torsion.

The troy weights are usually made in sets nested, each weight, therefore, being a cup inclosing all the weights of smaller denominations than itself, the smallest, usually 4 troy ounce, being solid, and fitting in a cup-shaped weight of the same denomination (Fig. 18).

As the avoirdupois weights, so the metric are composed of iron or brass; fine analytic metric weights are sometimes gold-plated brass and are sometimes platinum coated.

Fig. 17.-Brass block weights.

Fig. 18.-Troy cup weights (solid brass).

In purchasing avoirdupois and metric weights it is usually wisest to choose such as are arranged in a block. A very convenient form of such block is made of cast iron, the objection to wooden blocks being that they frequently swell, rendering it difficult to remove the weights.

The metric analytic weights should be in a fine wooden box, the depressions in which the weights rest being lined with velvet, serving the

double purpose of keeping the weights from being scratched and preventing them being held fast by the swelling of the wood. The denominations under one gramme are made of sheet platinum.

For dispensing purposes troy weights are used, the denominations under ten grains being made of aluminum, either in sheets or in the form of wire, the latter being bent into as many sides as the weights have grains.

MEASURES

In pharmacy the larger measures, gallons, quarts, etc., are not often used, although reliable measures of these sizes should be found in every well-regulated pharmacy. These are made of brass, tinned iron, or enameled iron.

Indispensable in the prescription department are graduates or measures, from a pint down, made of glass, with the subdivisions of the unit marked on the sides (Fig. 19). The graduation of these measures is performed in two ways: in one the plunger used in making the utensils

out of the molten glass is graduated; thus the finished graduate is produced in one operation. More popular, however, are those in which the graduating is done by hand, the quantity of liquid being poured into the ungraduated vessel from a burette, and this quantity of liquid is carefully marked, the finishing of the graduate being done by the engraver.

The subdivisions of graduations are into drachms and minims in ordinary graduates, and into mils in the metric.

The minim graduate, however, is far from being reliable as a large amount of the liquid is retained by capillarity of the glass vessel, and, therefore, for measuring minute quantities, a pipette should be used. Such pipettes, graduated to minims, were put on the market by Dr. E. R. Squibb (Fig. 20), and are more reliable not merely by reason of the dispensing of the full quantity of the liquid, but, being narrower than a graduate of equal capacity, there is less chance for error in pouring more or less of the liquid into the utensil; the narrower the graduate, the more accurate can the quantity desired be determined.

In this way, for delicate analytic work, a liter graduate is rarely used,

Such

there being substituted for same a flask of such size that when filled with a liter of water the water rises within the narrow neck of the flask. liter flask is used chiefly in volumetric analysis, liquids being delivered

and measured from same by means of pipettes (Fig. 22) and by burettes (Fig. 23).

In both cases the liquid is measured within a narrow tube, hence based on the same principle as the liter bottle (Fig. 21).

Fig. 25.-Pouring into graduate.

In placing liquids in a comparatively small utensil, like a burette, it will be observed that, through capillarity, the liquid is attracted to the sides of the vessel, thus making an edge which it is sometimes difficult

to read; such an edge is called a "meniscus," and in reading the quantity of the liquid, the line representing the quantity should be exactly in the center of the meniscus (Fig. 24).

It is hardly necessary to state that in measuring liquids in a graduate the latter should be held perfectly level and the line representing the desired quantity even with the eye (Fig. 25).

BIBLIOGRAPHY

Weights. (History) J. Q. Adams, Report on Weights and Measures, 1821; Ellis, A.J.P., 2, 1830, 111; Arny, Am. Dr., 42, 1903, 250 and 282; Anon, Dr., Circ., 58, 1914, 111. (Grain) Lloyd, A.Ph.A., 43, 1895, 137. (Gallon) Mason, Am. Dr., 17, 1888, 2; Lyons, Am. Dr. 17, 1888, 41.

573.

Size of Drops.-Payne, Dr., Circ., 41, 1897, 125; Eschbaum, Nat. Dr., 39, 1909, 19. Balances.-Brandel and Kremers, The Balance.

Standard Pipette.-Traube, Ph. Zt., 54, 1909, 203.

Graduate Making.-Parrish, A. Ph.A., 10, 1862, 199; Pilé, A.Ph.A., 21, 1873, Interpolation.-Dayton Miller, Laboratory Physics, 1903, 57.

CHAPTER III

SPECIFIC GRAVITY

IN Chapter II care was taken to mention that 454.6 grains was the weight of a fluidounce of distilled water. Why distilled water? Would it not be the weight of a fluidounce of chloroform as well? Common sense teaches that " 'some liquids are heavier than others," as we put it -that chloroform is heavier than water. A fluidounce of chloroform weighs about 670 grains, against 454.6 grains, the weight of the same volume of water. Accordingly, we see that chloroform is nearly one and one-half times heavier than water. That gives the idea of specific gravity, the definition of which can be given as the relative weight of equal bulks of different bodies, adding thereto the provision that, for solids and liquids, the unit of specific gravity is distilled water.

Now, to go back to our chloroform example: the volume, the bulk, the capacity of that one-ounce graduate remains the same whether it contains water or chloroform. Therefore a fluidounce-be it of water or of chloroform-represents "equal bulk." The respective weights of a fluidounce of the two liquids is "relative weight," and, reducing to unity, the specific gravity is obtained.

Instead of the term specific gravity, the word "density" is the more happy choice, since the term "specific gravity" suggests the gravitating force of the earth, which has only indirect connection with the subject.

The estimation of the specific gravity of various substances is of value, first, as the indication of purity or strength of the substance, and, second, through the data thus afforded we can estimate the volume of the substance, while in a minor degree it is sometimes of service in the diagnosis. of disease.

The first application of density given above can be best explained by a simple illustration, thus: The average 93 per cent. sulphuric acid has a specific gravity of 1.83; in other words, is 1.83 times heavier than If this acid is mixed with water, the density of the liquid will vary to somewhere between 1.83 and 1.000, and, indeed varies, in proportion to the amount of water added.

In the same way official 92.3 per cent. alcohol has a specific gravity of

0.810, and is, therefore, 0.8 times as heavy as water. If such alcohol be mixed with water, the specific gravity of the liquid will rise to some point between 0.81 and 1.000, and, indeed, in proportion to the amount of water added; hence by taking the specific gravity of such liquid, we can estimate the exact strength of the liquid. Tables of this character have been carefully worked out, and are found in the pharmacopoeia.

The other two uses of specific gravity being of minor importance, will be discussed in the appropriate place.

Estimation of the density of a body differs according as the substance. is solid or liquid, and since, by the explanation given above, the specific gravity of liquids becomes simple, we will first discuss the estimation of the specific gravity of liquids; afterward, that of solids.

There are several ways of estimation of the specific gravity of liquids: (a) By use of pyknometer or specific gravity bottle; (b) with the hydrometer; (c) with the Mohr-Westphal balance; (d) with Lovi's beads; (e) with the Jolly spiral balance.

ESTIMATION OF SPECIFIC GRAVITY OF LIQUIDS

With the Pyknometer.-As mentioned above, the estimation of the specific gravity of a liquid is rendered simple because of the ease with which the weight of a bulk of given liquid can be compared to the weight of an equal bulk of water. Thus, in the simplest methods of estimation all that is necessary is to estimate the weight of a fluidounce of water in a graduate, then pour out the water, dry the graduate, pouring in the same measure of the liquid, the specific

Fig. 26.-Pyknometer.

gravity of which is desired, weighing same, and then comparing the weight of the fluidounce of water with the weight of the fluidounce of liquid. As many times heavier as is the weight of the fluidounce of the liquid than the weight of the fluidounce of water, so many times heavier is that liquid than water, which figure represents the specific gravity. The inaccuracies of this method of estimation are due to the same causes as any inaccuracy in the measurement of liquids by means of the graduate, as mentioned in the preceding chapter, viz., the uncertainty in reading an exact volume of a liquid when placed in as wide a vessel as a graduate; hence, in order to obtain accurate results, we generally weigh the liquid, the weight of which is to be compared to the weight of an equal bulk of water, in a flask called a pyknometer, two modifications of which are shown in Figs. 26 and 27.

The only advantage a pyknometer (or specific gravity bottle) has over the graduate method just given is its greater accuracy. By tilting a graduate in measuring it is very easy to get a little more or a little less