While in the interior of the figures of which we speak the grains are clear, distinctly terminated, although of very variable size, in the boundary the grains are as if half effaced, stretched, stained; for the most part, indeed, they have disappeared to make way for trains of matter which have replaced the granulation. Everything indicates that in these spaces, as in the penumbræ of spots, the photospheric matter is submitted to violent movements which have confused the granular elements. We have already referred to the paradox that the sun’s appearance can now be best studied without the eye applied to the telescope. This is what Dr. Janssen says on that point. “The photospheric network cannot be discovered by optical methods applied directly to the sun. In fact, to ascertain it from the plate, it is necessary to employ glasses which enabled us to embrace a certain extent of the photographic image. Then if the magnifying power is quite suitable, if the proof is quite pure, and especially if it has received rigorously the proper exposure, it will be seen that the granulation has not everywhere the same distinctness; that the parts consisting of well-formed grains appear as currents which circulate so as to circumscribe spaces where the phenomena present the aspect we have described. But to establish this fact, it is necessary to embrace a considerable portion of the solar disc, and it is this which it is impossible to realise when we look at the sun in a very powerful instrument, the field of which is, by the very fact of its power, very small. In these conditions we may very easily conclude that there exist portions where the granulation ceases to be distinct or even visible; but it is impossible to suppose that this fact is connected with a general system.” But it is not alone with the uneclipsed sun that the new method enables us to make discoveries. The extreme importance of photography in reference to eclipse observations cannot be over estimated. Most of our best observations of eclipses have been wrought by means of photography. The time of an eclipse is an exciting time to astronomers; and it is important that we should have some mechanical operation which should not fail to record it. [Illustration: FIG. 212.—Copy of Photograph taken during the Eclipse of 1869. ]