solutions of ferrous sulphate or of stannous chloride containing some citric or tartaric acid. These sub-salts of iron and tin very rapidly absorb oxygen from the air or from any chemical substance which parts with it readily. With these solutions Stokes replaced the tissues. He abstracted the oxygen from the oxyhæmoglobin; then, shaking the solution of reduced hæmoglobin with air, he reproduced the action which occurs in the lungs. If the hand be held between a spectroscope and the source of light, in such a position that the beam passes through the thin tissue of two fingers where they are in contact, the spectrum of oxyhæmoglobin is obtained. If now the circulation through the fingers be impeded by putting strong indiarubber bands round them, the blood becomes venous, and the two narrow bands of oxyhæmoglobin give place to the broad band of reduced hæmoglobin. Although very soluble, hæmoglobin may be obtained in crystals, the form of which varies in different animals. When obtained from human blood, the crystals are rhombic prisms; from the guinea-pig, tetrahedra; from the squirrel, hexagonal plates. Yet it is unlikely that the hæmoglobin of one animal differs chemically from that of another in any proper sense of the term. Probably the form of the crystals depends upon the amount of water of crystallization. The apparent polymorphism of hæmoglobin may be associated with the great size of its molecules (_cf._ p. 66). Even when in the crystalline form, hæmoglobin can take up oxygen; but the difficulties which attend its purification and crystallization render somewhat uncertain the amount of oxygen which a gramme of crystallized hæmoglobin can absorb. In solution, 1 gramme can take up 1·34 cubic centimetres. The whole of the hæmoglobin of the body would, therefore, if it were all in the oxidized condition, hold about 4 grammes of oxygen. It is not with oxygen alone that hæmoglobin can combine. It can absorb the same volume of carbonic oxide or of nitric oxide gas. Both of these gases it holds more firmly than oxygen. Neither carbonic oxide-hæmoglobin nor nitric oxide-hæmoglobin is of any use to the tissues. If the blood becomes charged with the fumes of carbonic oxide (CO) given off by a coke-fire, this gas proves extremely poisonous. The blood does not lose it in its circuit through the body, nor is it exchanged for oxygen in the lungs.