70 ON THE GENERAL PROBLEM OF PHOTOGRAPHIC REPRODUCTION [359 requires a more powerful source. Good results were obtained with. a limelight; the portrait, barely visible at all on the single plates, came cmt fairly well under this illumination. If it were proposed to push the experiment much further by the combination of a larger number of plates, it would probably be advantageous to immerse them in benzole contained in a tank, so as to obviate the numerous reflexions at the surfaces. It has been mentioned that in the above experiment the development of the plates was rather light. The question may be raised whether further development, or intensification, might not make one plate as good as two or three superposed. I think that to a certain extent this is so. "When in a recent experiment one of the plates above described was intensified with mercuric chloride followed by ferrous oxalate, the picture was certainly more apparent than before, when backed by a sufficiently strong light. A-iid the process of intensification may be repeated. But there is another point to be considered. In the illustrative experiment it was convenient to copy all the plates from the same negative. But this procedure would not be the. proper one in an attempt to render visible the solar corona. For this purpose a good many independent pictures should be combined, so as to eliminate .slight photographic defects. As in many physical measurements, when, it in desired to enhance the delicacy, the aim must be to separate feeble constant effects from chance disturbances. It may be that, besides that of the corona, there are other astronomical problems to which one or other of the methods above described, or a combination of both, might be applied with a prospect of attaining i\ further advance.