The heart of the frog, of the turtle, of the tortoise, and of the shark will beat regularly for a long time after removal from the body, if supplied with a regular stream of nutrient fluid. The fluids used for this purpose are the blood of the herbivora diluted with common salt solution, or a serum albumin solution, or a 2 per cent. solution of gum arabic in which red blood corpuscles are suspended. The simplest apparatus to use is that known as “Williams’.” Williams’ apparatus consists of two glass bulbs (see diagram), the one, P, containing nutrient fluid to which a known quantity of the poison has been added; the other, N, containing the same fluid but to which no poison has been added; these bulbs are connected by caoutchouc tubing to a three-way tube, T, and each piece of caoutchouc tubing has a pressure screw clip, V¹ and V; the three-way tube is connected with a wider tube containing a valve float, F, which gives free passage of fluid in one direction only, that is, in the direction of the arrow; this last wide tube is connected with a Y piece of tubing, which again is connected with the aorta of the heart under examination, the other leg of the Y tube is connected with another wide tube, X, having a float valve, F²: the float containing a drop of mercury and permitting (like the float valve F) passage in one direction only of fluid, it is obvious that if the clip communicating with N is opened and the clip communicating with P is closed, the normal fluid will circulate alone through the heart; if, on the other hand, the P clip is open and the N clip closed, the poisoned blood will alone feed the heart. It is also clear that by raising or depressing the bulbs, the circulating fluid can be delivered at any pressure, high or low. Should a bubble of air get into the tubes, it can be got rid of by removing the cork at S and bringing the fluid up to the level of the top of the aperture. The observation is made by first ascertaining the number and character of the beats when the normal fluid is circulating, and then afterwards when the normal is replaced by the poisoned fluid. A simpler but less accurate process is to pith two frogs, excise their respective hearts, and place the hearts in watch-glasses containing either serum or a solution of common salt (strength 0·75 per cent.); to the one heart is now added a solution of the poison under examination, and the difference in the behaviour and character of the beats noted. The phenomena to be specially looked for are the following:-- 1. The heart at the height of the poisoning is arrested in diastole. 2. The heart at the height of the poisoning is arrested in systole.