ON SENSORIAL VISION. 403 



on sensitive paper dies out, and can be replaced by 

 another without the renewed application of any chemical 

 agent. 



(4.) Thus considered, ocular spectra are quite as much 

 entitled to be considered as things actually seen, as the 

 retinal pictures of which they are the successors, or rather 

 remnants. It is quite otherwise with that other class of 

 visual impressions to which I now refer, and which differ 

 altogether from ocular spectra, not only in being for the 

 most part (though not always) much less vivid and much 

 more dreamy (if I may use the term without casting a 

 doubt on their reality as facts\ but also in having no 

 reference or resemblance to any objects recently seen, or 

 even recently thought of. Of course, when I speak of 

 their reality as facts, I do so on the ground of their 

 admitting of being watched and studied with the same 

 sort of wide-awake attention which might be given to 

 any faint and fugitively-presented real object : though it 

 is no more possible *to describe them accurately, much 

 less to draw them, than it would be to do so in the case 

 of objects dimly seen in the dusk of evening, and capri- 

 ciously appearing and disappearing. But this does not 

 preclude their being observed and described, pro tanto, 

 in general terms. 



(5.) I fancy it is no very uncommon thing for persons in 

 the dark, and with their eyes closed, to see, or seem to 

 see, faces or landscapes. I believe I am as little vision- 

 ary as most people, but the former case very frequently 

 happens to myself. The faces present themselves in- 

 voluntarily, are always shadowy and indistinct in outline 



