a proportion being retained as a mobile reserve in order to support the action of the offensive reserves in the divisional sectors of defence. Only a small proportion of the available machine-guns should be allotted for the support of the troops in the outpost zone. No attempt should be made to site machine-guns so that every yard of ground is swept by their fire, which should be reserved for protection on a larger scale, covering the more important features and denying to the enemy the most favourable routes of advance. Localities of tactical importance must be strongly covered even though it becomes necessary to leave gaps on parts of the front where an attack is considered less probable. Guns must be sited primarily for the defence of ground by direct fire, but, subject to limitations as regards expenditure of available ammunition, the long-range fire of guns in rearward positions should be utilized for the assistance of troops in the more forward areas.
Such were the tactical principles which were made possible of application by the organization of machine-guns on a divisional basis. Experience showed that a rigid battalion organization on the model of the infantry battalion was not the most suitable for machine-guns. The application of the tactical principles enunciated above necessitated the organization of the divisional machine-guns into groups of varying sizes. But as the machine-gun company was not self-contained as regards train transport and supply, the splitting-up of a company in order to organize the required groups led to considerable administrative difficulties. The organization of a battalion by grouping together the machine-gun companies already in the division was the quickest and least expensive method of placing the machine-gun organization on a divisional basis; but it was subsequently realized that a more suitable organization would have been attained by making each machine-gun company self-contained in all administrative matters in such a manner that sections could be detached as and when required. (J. S. HA.)
MACHINE TOOLS (see 27.21). The decade 1910-20 saw a noteworthy development in every branch of machine-tool engineering. In no branch was the progress more marked than in instruments for precise measurements. These include types em- ploying both physical and optical means. Their perfection has made possible the production of interchangeable parts in com- mercial quantities. Without means of accurate gauging the making of cheap automobiles in great numbers would be im- possible. This is also true of rifles, typewriters, sewing-machines and hundreds of other things made and used daily in great numbers. For accuracy and almost universal application, the gauge blocks shown on Plate I., fig. i, made by C. E. Johansson, Eskilstuna, Sweden, stand high. The first combination set on his system was made in 1897, but not until 1911 was Johansson able to produce them in commercial quantities of a guaranteed quality. Subsequently these blocks became so recognized as standard that there is hardly a manufacturing plant in the world doing accurate or interchangeable metal work that has not one or more sets for reference purposes or actual use. They are also in con- stant use at the National Physical Laboratory, London; the National Bureau of Standards, Washington; the Bureau Inter- national des Poids et Mesures, Paris, and similar institutions of all the principal nations.
A full set consists of 81 blocks with surfaces flat and parallel within one hundred-thousandth of an inch. A standard set is made up of four series. The first series consists of nine blocks, the first o-iooi in. wide, increasing by o-oooi in. each to the ninth, 0-1009 in- wide. The second series consists of 49 blocks, the first o-ioi in. wide, increasing by o-ooi in. each to the 49th, 0-149 in. wide. The third series consists of 19 blocks from 0-050 in. to 0-950 in. wide, each increasing by 0-050 in. The fourth series consists of four blocks I, 2, 3 and 4 in. wide respectively. These blocks may be stacked or " wrung " together to form an enormous number of very accurate " blocks " practically equal to a similar solid block. For instance, the blocks of the fourth series can be combined to give any size in even inches from one to ten. The blocks of the third series can be combined with those of the fourth so as to give any even multiple of 0-050 between 0-050 and 10 in. The second series furnishes means of stacking the gauges to obtain dimensions varying by thousandths, and the first series gives variations by ten-thousandths of an inch. One stack of all the blocks wrung together gives accurate results.
That these blocks are held together by far more than atmospheric pressure is proved by a demonstration given Nov. 10 1917 before the Stockholm Technical Institute. Two blocks were wrung together. The sizes of the two surfaces in contact were 0-49 sq. in. and they sustained a weight of 220 Ib. The atmospheric pressure contributed about 6-6 Ib., from which it will be seen that the adhesive power of the blocks was more than 30 times atmospheric pressure. In spite of this extraordinary adhesive power the blocks are easily separated by a simple sliding movement, and they are as easily "wrung " together in the same way if the surfaces are first wiped with the hand. The great advantage given by these blocks is that they furnish a prac- tically universal standard of gauging, since parts, gauges, templets or tools, made in England and checked with reference to them, will check the same with a set in America, France or Japan. The com- position of these gauging blocks is such that they are long-wearing and little affected by ordinary changes in temperature.
A gauge known as the Prestwich fluid micrometer is shown on Plate I., fig. 2. This is the invention of John Alfred Prest- wich, of the English firm of John A. Prestwich & Co., Ltd. It was originally developed about 1910 for use in his own works, but later was put on the market.
The gauge is shown with a piston ring between the gauging points. The lower gauging point, or " anvil," is a stationary block of hard steel set into the base of the instrument. The movable gauging point is set directly above the anvil and is attached to the lower side of a thin, springy diaphragm of metal which forms the bottom of a fluid container about 2\ or 3 in. in diameter, and about J in. thick. A small glass tube leads upward from this container and a coloured liquid is put into the container and extends part way up into the glass tube. Pressure on the movable gauging point presses the diaphragm upward and causes the coloured liquid to rise higher in the glass tube, where it is plainly visible. A graduated scale at one side of the tube shows the amount of upward movement, and pointers at the left of the scale may be set to show the limits for various kinds of work. Owing to the size of the diaphragm and the small hole in the tube, any movement of the gauging point is greatly multiplied by the liquid in the tube, and in some of the instruments a variation of one thousandth of an inch between the gauging points will cause a difference of half an inch in the height of the liquid in the glass tube. This instrument is especially valuable for quickly inspect- ing machine products, as any variation is instantly visible.
During the World War considerable difficulty was experienced in finding a satisfactory method of quickly inspecting screw threads for size, shape and lead. This was solved in 1916 by the National Physical Laboratory in England, under the direction of Sir R. T. Glazebrook, by means of a projection lantern. The general principle of this lantern is along the lines of the stereop- ticon or the motion-picture machine, as the threaded (screwed) piece to be inspected is placed in the path of a powerful beam of light which projects a greatly enlarged image of the object upon a screen. An accurate drawing of the screw is previously imprinted on the screen, and the faithfulness with which the pro- jected image conforms to the lines of the drawing instantly de- termines the accuracy or inaccuracy of the screw or any part of it. Building on this original idea, a number of concerns have placed on the market " comparators," " projectoscopes," " pro- jection lanterns " and similar instruments under various names.
A measuring machine sufficiently accurate for all ordinary shop purposes is illustrated on Plate I., fig. 3. It consists of a bed with a sliding work-table and a microscope mounted on a com- pound slide. The latter is furnished with a large dial micrometer reading to o-oooi in. In addition the microscope is fitted with two hair-lines, one rotating with the eyepiece and one with the outside tube. This is fitted with a dial reading to half degrees, while the eyepiece carries a vernier reading to one minute.
The accuracy of measurements depends upon standard rods inserted between blocks at the left-hand end, one block being on the bed and the other on the work-table. The work-table is provided with centres, one of which has cross adjustment for alignment. The method of using this machine will be evident in the case of limit gauges and the like, having plain length measurements, since the selection of suitable measuring rods, the setting of the hair-lines and the reading of the traversing micrometers present no difficulties. With little additional trouble the machine may be used for contour work, while the two hair-lines enable the operator to measure the pitch and angle of screw threads as well as the depth. The table will accommodate work 12 in. long and up to 3 in. in diameter. A lamp and mirror are set as shown in the illustration to give clear projection.
For testing the flatness of a lapped steel surface of a gauge, or other polished surface, the U.S. Bureau of Standards has developed
