Instead of having a circle divided into four equal parts of 90 deg. each, we have the first of 100 deg., the second of 90 deg., the third of 90 deg., and the fourth of 80 deg., or some similar arrangement. This, of course, lacks much of the ideal symmetry which is required not only by theory, but rigorously demanded by the results of practice.

To overcome this lack of symmetry and also that due to the fact that the total number of barrows in a charge is frequently not divisible by the number of dumping points around the hopper, we have a simple means not as universally used as it should be. This consists in shifting the first point of dumping on the bell one space each time. A very easy method of accomplishing this result is to have a number of barrows, one more or one less than an even multiple of the number of barrows in the charge. Then if pains be taken to see that the fillers go round the top always in the same direction, the first barrow dumped on each charge will always be one space away from where it was the last charge. On the next, another space, and so on until within a comparatively few charges the circuit has been made, and the first barrow comes back to its starting point. This means that while each individual charge lacks much of symmetry, its initial point is revolved so that each portion of the furnace in turn receives the same treatment, thus producing what is known as "spiral filling," since any given point, such as the dumping point of the first barrow, in the suc-cesssive charges as they descend through the furnace, makes a series of steps very similar to those of a "spiral" (helical) staircase.

How old this method of filling may be I do not know; it was first called to public attention in a paper before the American Institute of Mining Engineers by W. A. Barrows, then at Sharpsville, Pennsylvania, about ten years ago. On account of the important irregularities in any one charge whose origin I have tried briefly to describe, this or some similar system, giving analogous results, should always be used in hand-filling.

When two top-fillers are employed another system can be employed with good results. This consists in having the two men dump their barrows at points immediately opposite one another, and at the same instant. The next time they dump opposite to each other again, but in the case of 4-point dumping this, of course, means, at right angles to the first position. The two pairs of heaps so produced are not similar but each one is similar to its opposite, and by alternating the position for the first dump the series of charges formed are all symmetrical about two axes, although the heaps on one axis are not like those on another, but these are alternated so that in the course of a double charge we even up the inequalities of one charge with those of the next. It is more difficult to secure this sort of filling than it is spiral filling, and this kind should not be used in preference to the latter except under unusual circumstances.

The size of the charge unit to be used, and whether the coke and ore are to be dumped all at once or separately are questions of operation rather than of construction, and will therefore not be discussed here. But in a general way we may say that coke charges range from six thousand pounds for small furnaces, up to fifteen thousand, and even eighteen thousand pounds on larger ones, while the best practice undoubtedly favors dumping the coke separately from the ore and stone. Similarly the question of the diameter of the bell which controls the radial distribution of the stock is a question of operation, and not one of construction, since the same construction is used in both cases, with only a relative slight difference in dimensions, and this question can therefore be passed over here.

Fig. 28. The Bauman bell and hopper.

Fig. 29. Fimstone's modification of the Bauman bell and hopper.