The hydrometer was never intended to be used with semi-solid substances such as tomato pulp, but only with liquids such as brine, sugar syrup, etc., in which there is no solid matter in suspension. With liquids, the number of degrees registered on the hydrometer when it is immersed in the liquid is equal to a definite specific gravity. This is not true with semi-solids such as tomato pulp, as each packer must work out the relation between the degrees of the hydrometer and the specific gravity on pulp under the conditions which obtain in his plant. This is a very simple thing to do, and once the packer has established this relation, it will hold good as long as he is in the pulp packing business. The reason why this relation is not the same for all packers is because tomato juice is not screened to the same degrees of fineness in all plants, and because all hydrometers having the same scale will not work the same on pulp. This is because they have different shapes. The shapes may only vary slightly, which will not interfere with their accuracy on clear liquids, but it will make a difference when the hydrometer is used in tomato pulp. Furthermore, some of the hydrometers will probably have too much variation in their diameter at different points, which is a disadvantage when they are used in tomato pulp. I have therefore always had the hydrometers which I have used for this test made to order and to conform to certain specifications which I laid down. [Illustration: Fig. 3. Special hydrometer for testing tomato pulp and catsup. (C. J. Tagliabue Mfg. Co.) ] In Fig. 3 is illustrated a hydrometer which is manufactured by the C. J. Tagliabue Mfg. Co. of Brooklyn, and which is made to conform to these specifications, which are: first, perfect balance; second, a minimum amount of variation in the diameter of the various parts of the spindle with the slope very gradual; third, a Beaume scale reading from 0 to 50 degrees; fourth, that it can be obtained in exact duplicate in any quantity. If the diameter at various points in the spindle is slight, with the slope very gradual, it seems to be an advantage. It is, of course, necessary that exact duplicates can be obtained to use when the first one, on which your calculations are based, is broken. In order to avoid disappointments which would be very apt to result from the use of a hydrometer which was not particularly adapted to this test, and which would very likely be difficult of exact duplication, the type illustrated above should be adhered to, and it will be found to give good results when used as directed in this chapter. Two pulps of the same specific gravity, and of practically the same degree of fineness, if tested hot, will register the same on the same hydrometer, or on two hydrometers which are exactly alike in every respect. (The hydrometer does not work well on cold pulp unless the pulp is thin.)