Many modifications of the bell have been made with the object of securing better distribution and better furnace work. These have sometimes been only changes in dimensions, but many other variations have been used as well. A number of these are shown in Figs. 25, 28, 29, 30, 31, 32, 33, taken from a paper by the late T. F. Witherbee before the American Institute of Mining Engineers, Vol. XXXV, entitled "Special Forms of Blast-Furnace Charging Apparatus".

Fig. 31 is representative of a number of these, this being the Firm-stone modification of the Bauman or double bell. The center of the cone forming the ordinary bell is cut away by an inverted cone having the same angle of slope as the bell, normally 45 deg., and the opening so formed is closed by a (nearly) stationary part commonly called, by those who use this type of top, the "plug." This is not quite stationary for the reason that the inner and outer surfaces of the bell must both make a good joint. It would be impossible to maintain an adjustment of the parts which would secure this if both seats were rigid, so the plug is counter weighted until its effective weight is very small and allowed a drop of about a half or three quarters of an inch, which permits it to follow the bell down for this distance on the opening movement, while on the closing movement the bell picks it up and makes the joint on its inner seat first, then reaches its seat on the outer surface or hopper afterwards, carrying with it the plug through the distance the latter dropped in opening.

Fig. 30. The Durham bell and hopper.

Fig. 31 differs from the other types shown in that the gas outlet takes place through its center, there being no flues in the side walls of the furnace. This secures an absolutely symmetrical distribution of the gas as well as of the stock, and while the former is less important than the latter, it is still very important.

Fig. 31. The Bauman-Firmstone bell and hopper at the Long-dale furnaces.

Tight joints in the pipe are maintained in spite of the rise and fall of the center plug by ball joints at each end of the horizontal pipe, shown broken off at the left-hand side of Fig. 31. The ball joint not shown is at the top of the down-corner over the side of the furnace, while the one in the vertical center pipe is shown at the bottom of the "tee" which forms the junction between the vertical and horizontal pipes. The small amount of vertical movement necessary in the center pipe is taken up by these ball joints without any difficulty. The center pipe, with the plug fastened to it, is suspended from the tee by turn buckles, of which the one on the near side is clearly shown in the illustration.

Fig. 32. Bauman double bell and hopper at the Saucon furnace, Hellertown, Pa.

In this type of top the lever which supports the bell can of course not be located in the center, and is therefore made double and takes hold of the bell at four points. See Fig. 34, which shows the construction used in Fig. 31 also. This type of top was used at the plant of the Longdale Iron Company practically throughout its entire life, and combined with careful hand-filling gave excellent results. The great advantage of the double bell, over the single is that there are two openings of different diameters, which combined with the reversed slopes inside and outside the bell, permit the stock to be distributed more advantageously in the furnace than is possible under most conditions with the single bell. The latter is confined to a maximum of three annular areas of different sizes of material, and may give only two, whereas with the double bell there can be two rings of fines, and three of coarse material, thus permitting a better distribution of the gases over the whole area of the furnace, and exposing the stock to this influence much more effectively.

Fig. 33. The Witherbee double bell and hopper.

This top has been used at many other plants beside Longdate, and has been very generally commended by those who have used it, but it is undoubtedly somewhat more complicated than the single bell top, and that, with the extreme difficulty of proving the advantage of one or another method of distribution on the furnace unless one of those under consideration be very bad, has prevented its general adoption.

A still more radical departure from the standard single bell top is the Langen charger in which a central member like a "plug" seats downward in the bottom of the hopper to close it, instead of being pulled up from beneath, as in the standard type. The "bell," which is still so called in this case, is raised, to open it, instead of being lowered as is the regular bell. Frank Firmstone developed the double Langen charger in which the discharge took place around circles of two different diameters as in the Longdate double bell, but both openings were produced by raising the closing-pieces or bells, first one and then the other. He has described this top as applied by him at the plant of the Glendon Iron Company in the transactions of the American Institute of Mining Engineers, Vol, XIII.

Fig. 34. Double Langen charger designed by Frank Firmstone for the Glendon Iron Works.

The top is well shown by Fig. 34, reproduced from this paper.

Excellent results were secured with these bells, but more care and skill was necessary in designing and operating them than with the ordinary single bell, and they have never had ft wide use in this country, although bells of this type are still extensively used in Germany.

There is one important difference in the conditions under which these tops had their principal use and those which predominate to-day. This is in the size of the ore, which in those times was almost entirely lump comparable in size with the fuel, while now the ore is almost exclusively fine.

In order to have the gas travel uniformly through the stock it is very important to have the greater portion of the fines next to the walls, since with material of uniform size the spaces next to the wall are about twice as great in proportion as they are in the body of the material, because when the high points of the individual pieces interlock with one another as they do in the mass of the material, the interstices are very largley filled, but when the high points come against a flat surface the interstices are preserved complete.

By taking spheres we can determine the amount of the interstices under each condition. We find that if there be a layer of spheres of uniform size arranged in close contact in equilateral triangles (the closest possible spacing) with another layer of like spheres resting in the low points of the first as they would arrange themselves naturally, the voids around the half spheres adjacent to the wall would be 65 per cent. of their volume, while those around the other half (away from the wall) would be 35 per cent.

This shows quantitatively the effect of the wall in increasing the interstices under conditions in which we can calculate it, and the general nature of the effect is the same in irregular materials.

It is an old tradition in the business that the fine material must be put against the walls, and it is obvious from these considerations that this is correct, and therefore the fines, that is the ore in modern practice, must go largely against the wall if we are to avoid excessive flow of gas there. Therefore we need now a single bell of rather large diameter in proportion to the stock line, but in the days of lump ores it was possible to carry this to an extreme and to have the charge column too open through the center and an excess flow of gas there, so it was desirable under those conditions to dump part of the charge in the center so that some fines would reach that part and help to obstruct it.