As a rule, it is very difficult to detect leucocytes in the act of eating red corpuscles; but under various circumstances their activity in this respect may be stimulated to such a degree as to show them, in a microscopic preparation, busily engaged in this operation. The writer had the good fortune to prepare a spleen which proved to be peculiarly suitable for this observation (Fig. 5). His method was an example of the way in which a physiological experiment ought not to be conducted. Having placed a cannula in the aorta of a rabbit, just killed with chloroform, he was proceeding to wash the blood out of its bloodvessels with a stream of warm normal saline solution, when the bottle from which the salt-solution was flowing overturned. Fearing lest an air-bubble should enter the cannula, he hastily poured warm water into the pressure-bottle, and threw in some salt, in the hope that it would make a solution of about 0·9 per cent. The salt-solution was allowed to run through the bloodvessel for rather more than an hour. When sections of the spleen were cut, after suitable hardening, every section was found to be packed with leucocytes gorged with red corpuscles. Some of the corpuscles had just been ingested; from others the hæmoglobin had already been removed. It may be that, for some unknown reason, the destruction of red corpuscles was occurring in this particular rabbit with unusual rapidity at the time when it was killed; but it seems more probable that the animal’s leucocytes were provoked to excessive activity by changes in the red corpuscles brought about by salt-solution which was either more or less than “toxic.” As a score of attempts to reproduce the experiment, with solutions of different strengths, have failed, it is impossible to be sure that this is a valid explanation. There must be something in the condition of worn-out red corpuscles which either makes them peculiarly attractive to predatory leucocytes or renders them an exceptionally easy prey. It does not require much imagination to picture the drama which is enacted in the spleen. Slow-moving leucocytes are feeling for their food. The majority of red corpuscles pass by them; a few are held back. The leucocytes, like children in a cake-shop, cannot consume all the buns. A selection must be made, and preference is given to the sticky, sugary ones. Red corpuscles when out of order show a tendency to stick together. When blood is stagnating in a vein, or lying on a glass slide in a layer thin enough for microscopic examination, its red discs are seen after a time to adhere together in rouleaux. The parable of a child in a cake-shop is not so fanciful as it may appear.