Page:The third Huxley lecture.pdf/21

17 leave a deep layer of serum. In whipped cow's blood, rouleaux forming in the serum as in the plasma, there will be found, if the animal was healthy, only a thin superficial serous layer, even after the lapse of 24 hours.

I once drew blood from a donkey into two similar glass vessels, one empty, the other half full of water. The diluted blood and the undiluted clotted in exactly the same time. But whereas in the normal blood there happened to be an unusually thick layer of buff, comprising nearly two-thirds of the whole mass, the watered blood gave no buff, and the microscope showed that the red corpuscles had lost their natural adhesiveness.

Human blood, as is well known, shows the buffy coat in some states of inflammation. But it may also occur in anæmia. And it may well make our profession humble to reflect that in days within living memory buffing of the clot was regarded as an indication for further withdrawal of the vital fluid by venesection.

To return from this digression: adhesiveness of the corpuscles, both red and white, was seen in the vessels of an irritated area of the frog's web, as in blood outside the body. But in a perfectly healthy part no such condition was observed. A string tied round a frog's thigh of course made the blood in the vessels of the foot motionless; but on the slightest