1. Injection of pituitary extract has an immediate action on milk secretion, but the effect soon passes off. During the period after injection there is a rather rapid decrease in the milk‐flow, but this decrease is gradual and there is no sudden drop followed by increasing secretion to normal. This indicates that the effect of the extract is not muscular.2. From the result of successive injections at various intervals as well as from numerous other observations it is also concluded that the effect is not muscular.3. The daily yield is only slightly increased as a result of injection. This shows that the extract acts rather by setting free than by causing the formation of the milk constituents.4. From the composition of the milk obtained as a result of injection, together with other facts, it is concluded that the action of the pituitary extract is not effected through rise of blood‐pressure.5. The amount of milk secreted as a result of injection depends on the period of lactation and also on the state of nutrition. This together with other considerations supports the view that it is not variations in the condition of the pituitary gland which give rise to the cycle of changes which occur in the mammary gland.6. Histological evidence points to a direct action of the extract on the glandular epithelium. The flattened appearance of the cells of the full alveolus is caused by the expansion which they have to undergo when the lumen becomes filled with milk.7. The milk obtained as a result of injection is normal in composition except for a higher percentage of fat; in the following milkings, however, there is a drop in the percentage fat although that of the other constituents remains normal.8. While the solids‐not‐fat (proteins, lactose, and ash) are secreted in close connexion with the water of the milk, the amount of fat secreted is in no wise connected with the amount of water. The ratio “nitrogen to lactose” is relatively constant throughout.9. These facts suggest the theory that the pituitary extract causes the combination of the precursor of milk‐protein and lactose (possibly a glyco‐protein) with water and salts of the blood, and so by a purely secretory action produces this part of the milk. The suddenly altered tensions so set up in the epithelial cells cause the fat globules which have accumulated at the ends of the cells to be discharged and to produce milk which is rich in fat. It is because of the relatively large amount of the pituitary extract injected and because of the suddenness of the change that the fat percentage goes up. The act of milking, which also causes rapidly altered tensions, has the same effect.10. Although there is considerable variation, yet the average results on varying the dose of pituitary extract injected show that the smaller doses give less milk with a lower percentage fat. When the doses are relatively large it does not matter how much is injected, for only a certain amount of milk can be obtained. There is some indication that a goat in the early stages of lactation is more sensitive to small doses than one in a later stage.11. Experiments with fractional milkings show that while the percentage fat rises from start to finish in normal milk, yet in milk obtained as a result of injection the percentage remains constant throughout.The expenses of the investigation were largely defrayed by a grant made by the Board of Agriculture to Dr F. H. A. Marshall.