306 H. M. EICRNARD. 



Ramon j Cajal, seemed to clear up the mystery. According 

 to the results obtained by these methods, tlie fibre of tlie 

 optic nerve still runs into the ^'ganglionic cell" as shown in 

 the old diagram, but there is no longer any necessity to trace 

 any single one of the distal processes of the cell thi'ough the 

 retina ; for direct fibrillar connection with the rods is not 

 required. The new docti'ine is that the ramifications of the 

 ganglionic cell within the inner reticular layer are able to 

 receive the stimulus from the rods through the mediation of 

 opposed ramifications of other more distal cells by some kind 

 of intermittent contact. 



One of the very earliest results of my own study of the 

 retina, however, was the conviction that the thick cyto- 

 plasmic axis-cylinder process joining the ganglionic 

 cell to the optic nerve has no existence. Though 

 during the last ten yeai'S I have examined many hundreds of 

 preparations of vertebrate retinas fixed and stained by all the 

 best known methods which are relied upon for ordinary 

 histological work, I have never once been able to see it. 

 In sections cut along the optic nerve it surely ought to have 

 been seen, especially when the distal processes of the same 

 cell were always visible. It was long before I could bring 

 myself to deny the existence of so important an element of 

 the above-mentioned familiar diagram, especially in the face 

 of the direct support which the recent work of Ramon y Cajal, 

 Dogiel, and others appears to lend to it. And, indeed, I did 

 not do so until I had something positive to put in its place. 

 It is now, however, perfectly certain that the appearances 

 which gave rise to the diagram, and still seem to give rise to 

 it, have been wrongly interpreted. 



Beginning with the observation which first shook my faith 

 in the correctness of the old diagram, 1 found that, if we take 

 a survey of the retinas of a number of different vertebrates, 

 only a small proportion of the ganglionic " cells " have any 

 even superficial resemblance to the diagram. It has been 

 constructed almost entirely upon phenomena most frequently 

 found in the Mammalia, and rarely in the lower forms. In 



