place, as from its thinness it is apt to deflect from the straight line. The slide will then be free to move easily backwards or forwards, and, holding the brass cock between the finger, and thumb of the right hand, bring it to the ends of the staves near your left hand. It is a matter that varies in different casks as to what point in the staves to take; the rule is, to take as nearly as possible that part which from its position must be opposite the angle of the chimb inside the cask (in the one we are gauging this will be nearly the outside of the stave). When the rule is thus adjusted to the proper position, grasp it more firmly in the left hand to fix the slide; remove it from the cask, and read the diameter on the line of inches on the lower part of the face of the rule, at the point opposite the brass cock. The back head must be first taken this way, and the dimension marked on the right of the bunghole as you stand when taking it. Then, coming again round, take the front head, and compare with the back head, as marked on the bouge. Suppose the back head in the present case 22.8, and the front head 22.6, the mean will be 22.7, which mark on the head of the cask between the length and bung. Our dimensions will now stand thus: length 49.7, head 22.7, bung 31.3; and the head rod, having just been used to take the head diameter, will be convenient for the next operation, which is casting out the content. To do this, adjust the cock to the mean head diameter thus found, and look along the lower line of the rule for the bung diameter;—exactly over this on the middle or spheroidal line of the slide will be found 6; look for this on the line of inches on the slide below it, and immediately under, on the inch line of the rule, will be found the mean cylindrical bung28.7 inches, which is, as was explained before, the diameter of a cylinder that, at the same length as the cask, would contain the same quantity of liquor. Now set the gauge point which is on the upper line of the slide to the length 49.7 on the upper line of the rule, and look along the upper line of the slide for the mean cylindrical bung, above which, on the rule, will be found the content of the cask, which in the present instance will be 116 gallons. You have now obtained the content of the cask;

but if it have liquor in, and be not quite full, it will require another process to find the ullage. Set the brass on the bung rod to the mean bung as first found, namely 31.3, and having wiped the rule dry on the inch line for a short distance near the top, re-insert it in the cask until the brass touches the bungstave, taking care that it does not slip; then, quickly withdrawing it, note where the liquor has marked the rule, and write down this height on the head of the cask under the bung diameter. Now turn to the back of the head rod, and bring the bung diameter 31.3 on the lower line of the slide over 100 on the segment line, noting what division on the segment line is under the ullage on the lower line of the slide. This found, place the slide so that the corresponding division on it shall be over 100 on the segment line, and then look for the content of the cask on the upper line of the rule, under which, on the slide, will be the ullage required. When the dimensions of casks are smaller than the head rod is calculated for, they must be doubled, and the content thus found reduced afterwards. If the mean bung be doubled, it will be equivalent to doubling the figure in two dimensions (breadth and depth), and therefore the content will be four times more than required; if the length be doubled—as but one dimension is involved, the content is only doubled,—while if both length and mean bung are doubled, the result is 8 times more. So, when the dimensions are too large; if the mean bung be halved, the content found will be one fourth part of the truth; if the length be halved, the content will be halved also ; and if both dimensions be halved, the content will be one eighth of the actual content. [See chapter on ratios of figures, page 17.]

Were casks all of similar figure, and of known and uniform thickness throughout, there would be little or no difficulty in arriving at their true contents, and the above formula would answer for all sizes. As it is, however, not only are casks never true mathematical figures, being at best only approximations, but even as such they vary excessively, great differences often existing between casks of the same manufacture and dimensions, although the original intention may have been to make them exactly alike.

In the first place, the wood used by coopers is not sawn, but rent or hewn into shape, in order that the fibre may retain its utmost strength. The thickness is, therefore, never uniform. Again: the outlines of the staves, and hence the figure of the cask, are produced under the guidance of the eye alone; and although coopers by long practice become extraordinarily expert, it is evident that the line of curvature so formed cannot possibly be mathematically correct, or even in its own form invariable. It therefore not unfrequently happens that a cask may have a fine, full figure on one side, and a spare and bad one on the other, so that it would be likely to be either under or over gauged, according to the portion that may be uppermost; this inequality of figure being rarely perceptible, unless the cask be placed upright on its head and viewed from all sides. Again: a cask whose outward figure shall betoken every good point may, from inequality of thickness within, be nearly the reverse, as, in making up the casks, the staves are often thinned at the bouge to facilitate their curvature, and at the chimb to equalize them and improve the outward appearance, while the internal thickness and inequalities are left untouched elsewhere; on the contrary, some casks are purposely thinned at the quarter, and at the bung and chimb are left of their full thickness. It is evident, therefore, that the thickness of a cask is rarely to be learnt from an examination of it either at the bunghole or at the chimb, and occasional recourse must be had to experimental proof, by boring a hole in certain parts of the cask, and testing the thickness by a graduated wire. But even this is not always to be depended on ; for as the adjacent staves vary in thickness, sometimes from one-tenth to half an inch, it is evident that neither would be the true criterion for the whole. A further check is therefore occasionally necessary, and that is the actual proof by measurement in the copper cans, and sometimes the removal of the head of a cask and an examination of its general internal appearance; and a gauger should never lose any opportunity that presents itself of seeing the interior of a cask, or its actual measurement, and comparing results with his general mode of practice. It must not, how

ever, be presumed that these various difficulties are insuperable; nor must the careless gauger excuse himself by the uncertainties that exist. The observant practical man can bring his practice very closely to the actuality. But though at times the best gauger may fail, the cause will most probably be from some unforeseen difficulty, rather than from the habit of taking everything for granted.

There is a certain family likeness continually existing between casks from the same countries and provinces. The art of coopering is an ancient craft, and certain empirical rules and directions are faithfully transmitted from father to son, from master to man,—each country having its peculiar mode of proceeding to work, and its time-worn, time-honoured, ideal or model of perfection. This, therefore, to some extent simplifies the matter, as, from the similarity of make, the casks of any one particular country may be gauged by a similar mode of treatment within a certain range; liable, of course, to such variations as the necessity of any particular instance may require. We shall therefore give, further on, a brief summary of some of the principal varieties of imported casks and their peculiar modes of treatment. It is not at all necessary, in practical gauging, to treat of casks as to their relation to the four varieties previously described, not only from the impossibility of their being made exactly to the model, but from the fact that in practice the third variety, or that which may be considered nearest to it, is rarely, and the fourth never met with.

The relative proportions between the four varieties, supposing their dimensions taken for gauging to be the same, are respectively as 116, 111.9, 107-2, 103.3. Now the intervals between these varieties are very wide, being an average of nearly 4 gallons in each instance, while no rule is laid down for their calculation. Some other means, therefore, must be devised to meet the case. The best rule in practice is to consider every cask with reference to the spheroidal figure, and after making such allowances as its thickness may render necessary, to consider how far its want of figure detracts from its capacity; an addi

tional allowance to this amount must then be made, and the dimensions thus found, if correctly taken, will be those of a spheroidal cask of the same content, and can then be cast out by the head rod, which is only calculated for the first variety. It is in this that the intelligence and observation of a gauger are apparent; and to do this correctly must be the sole aim of his study.

The allowances for figure may be made on any one dimension, or upon all; and a reference to the figure, Plate IV., will shew the apparent effect of allowance on the length, the result being an imaginary figure between the one under operation and the spheroidal model. In practice it is generally most convenient to make the allowances by deductions from the length, each of the other dimensions being taken correctly as if the figure were perfect; but on this question each gauger may think for himself, though, as the casks at different periods often pass through different gaugers' hands in regauging, blending, racking, &c., a uniform mode is preferable, and less likely to lead to confusion.

Explanation of Plate IV.—AGBDHC, as expressed by an ellipsis, represents a section of a spheroidal cask. The same expressed by straight lines is one of the 4th variety with the same dimensions. The rules being calculated for the first, all the dimensions would be taken entire; but for the other, to allow for the difference between the figures, a portion of the length is deducted, and an imaginary spheroidal cask of shorter length, but equal content, is produced. The length of the imaginary cask is ef, that of the other being EF, and the whole is shown by the figure aGbdHc. The faint lines complete the outlines of the spheroids. WW is the surface of the liquor or wet inches; the difference of vacancy in each figure is readily discernible. GC is the diagonal, and is the same for each cask, though the difference in content is great.

The following few general directions may be taken for a guide in the matter of figure :

Casks with a very full quarter, in which the diameter on each

I