the later trials the higher acidity gelatin was used. The count of the molds was not normal, due to the frequent stirrings, which prevented spore formation, besides injuring the hyphae. ----- Footnote 4: The Journal of Infectious Diseases. 1909. Vol. 6, p. 393. The results varied, some pulps giving a much higher initial count than others, but they all agreed in having a comparatively slight increase in the first three hours, the large numbers which one is led to expect not being present until the pulp had stood for at least five hours and under the most favorable conditions; usually it requires a longer time. The plates and the direct count agreeing in the general trend, though the numbers obtained by the two methods varied. In the pulp obtained from the steamed tomatoes, the initial count was much lower in the tomatoes steamed eight minutes, being only 20 per cc in the plates, but the same thing was true of these in that the increase was very slow at first. The figures from all the trials, both raw and steamed pulp, and from the plates and direct counts, show that the theoretical estimation of the increase of organisms from the classic twenty minutes required for reproduction of an organism with the consequent progression, irrespective of the condition of the organism at the start, or its environment, will have to be modified. In the plates all colonies, aside from the molds, were counted as bacteria, but this would not give a very large error, as yeast does not reproduce at the same rate as do bacteria. The state of comminution of the product determines to a considerable extent the number of organisms which may be counted. The more finely the comminution, the greater the number. Two pulps made from the same material, one run through an ordinary cyclone and the other through a finishing machine, will show from 50 to 100 per cent more in the latter. Coarse pulp and coarse ketchup may be inferior articles and yet give the better results by the direct method. The effect on the mold is even more marked—filaments and clumps will be torn into many small particles. The total quantity is not increased, but it is distributed more nearly perfectly and thus occurs in more fields. In work done on meat to determine the technique which should be employed in the bacteriological analysis, comparison was made between shaking the sample and grinding it in a mortar with sand. In the three samples reported, the shaking gave only 3, 12, and 13 per cent, respectively, of those obtained from grinding.[5] ----- Footnote 5: Weinzirl, John and Newton, E. B. American Journal of Public Health. Vol. IV, No. 5. A finely comminuted pulp was vigorously shaken for definite times and samples taken as quickly as possible after the tenth, fiftieth, one hundredth, and two hundredth times shaken. The results were as follows: