infinitely small ones like the microbes pass. Although, the thickness of the calcareous mass here exceeds eight hundred feet, the filter is not homogeneous, but is fissured; and through these faults, these cracks, in which the circulation is yet slow enough for the water, coming in muddy at the level of the plateau, to issue at the spring with admirable purity, the microbes continue to percolate. It is true that the microbes I have found are not pathogenic, but the importance of the studies can nevertheless be comprehended. If common microbes, brought in by the waters that fall on the plateau, can be found eight hundred feet below it, there is nothing to prevent noxious microbes—those, for example, of typhoid fever, diphtheria, or cholera—which may live in the water from being found there. There is in this a very interesting problem of hygiene and public prophylaxis. We should, then, be suspicious of these beautiful crystalline springs when there are epidemics on the plateau from which they come. They may contain micro-organisms—some indifferent, others dangerous. The microbes which I found belong to the genus micrococcus. Two of them (Micrococcus aurentiacus and M. citreus) developed in fine colonies of orange and citron-yellow colors; a third (M. aquatilis) gave no coloring matter. I obtained microbic colonies after the twentieth hour.
What, in short, is this underground river Midroï (Fig. 2)? A large fissure through which flow the waters drawn from the plateau by the avens, orifices, and cavities of every kind which make an enormous sponge of the mass of the Gausses. In past centuries, when the mass of water that fell on the Gausses was considerable, Midroï acted regularly and gradually enlarged the fissure; but now it acts only intermittently. Its vent then affords it sufficient outlet. Let me speak of this vent, the fine spring of Rochemale, which issues from the rock a little more than one hundred yards west of the orifice of Midroï. By this narrow fault, the communication of which with Midroï is highly probable, although it is not demonstrated, two hundred and twenty thousand litres of water escape every hour. In case the supply is doubled after great rains, the water, which can not escape by Rochemale in so large a quantity and in the same lapse of time, rises and fills all the meshes of the sponge. If the supply is increased again, the water flows into the river Midroï, which then comes into operation and the level of which may rise several yards, as is shown by the traces of wash-marks left by recent inundations on the walls of the river and which are shown in our photograph. There exists, in effect, in the very heart of the Gausses, a considerable and eminently variable reservoir of water; it is a real lake, and through the thousand fissures, through all the meshes of this interior region, flow the waters of the plateau,