130 PATHOGENIC BACTERIA. Experience shows that 1000 c.cm. of agar-agar rarely go through one paper, and I always expect when beginning the filtration to be compelled to boil the material which remains on the paper again, and pour it through a new filter. The formerly much-employed hot-water and gas-jet filters seem unnecessary. If properly prepared, the whole quantity will filter in from fifteen to thirty minutes. If made from beef-extract, the agar-agar almost always precipitates a considerable amount of meat-salts as it cools. This should be anticipated, but, so far as I can determine, cannot always be prevented. The amount is certainly lessened by making the bouillon first, filtering it cold, then adding the agar-agar, and dissolving and filtering it. The difficulty of filtering the agar-agar has led Pliigge and others to adopt a method of sedimentation. An in- genious apparatus for this purpose has lately been devised by Bleisch. The methods can be simplified by using a small pharmaceutical percolator, the bottom of which is closed by a rubber cork containing a tube which extends nearly to the top of the percolator and is attached to .Ľa rubber tube with a pinchcock below. The melted agar- agar is poured into this, and kept in the steam apparatus mntil the sedimentation is sufficient to allow clear fluid to be drawn from the top. As the clear agar-agar is drawn off the tube is pulled down through the rubber cork, and more drawn off until only the sediment is left. Agar-agar is dispensed in tubes like the gelatin and bouillon, sterilized by steam by the intermittent process, and after the last sterilization, before cooling, each tube is inclined against a slight elevation, so as to offer an ex- tensive flat surface for the culture. After the agar-agar jelly solidifies its contraction causes some water to collect at the lower part of the tube. This should not be removed, as it keeps the material moist, and also because it has a distinct influence upon the cha- racter of the growth of the bacteria.