24 H. M. BERNARD. 



done entirely at the Royal College of Science in the private 

 laboratory of Mr. Woodward, not only with his permission, 

 but not infrequently with his active co-operation. Professor 

 Howes has kindly allowed me to date the paper from the 

 College. 



The cones iu the eyes of frogs have hitherto been thought 

 to end in points not far below the level of the junction of 

 inner and outer limbs of the rods (see fig. 1, where a rod is 

 outlined for comparison). This, however, is not the case (see 

 P]. 3, fig, 2, a, h, d, f, g). The reason for this error, in 

 spite of all the splendid work which has been done on the 

 retina, is simple. As is well known, the rods and cones have 

 a strong tendency to divide at the junctions of inner and 

 outer limb (i. e. at the * in fig. 1) under the action of all 

 the ordinary reagents, hence it is hardly surprising that 

 the exquisitely delicate tips of the cones fail to survive 

 their violence. If, however, the fixative used be a sudden 

 short application of intense heat — five to ten seconds' 

 immersion in boiling corrosive acetic, the animal having 

 been kept and killed in the dark — the retinal elements, cones 

 as well as rods, will, in successful preparations, be found to 

 remain unbroken from the membrana limitans externa to the 

 pigment layer. 



Here it is necessary to state very distinctly that I am 

 referring only to the retinas of the frog and of some other 

 Amphibia. For while it is evident that the cones throughout 

 the Amphibia are analogous structures, certain differences, 

 e. g. in the positions of the nuclei, make it unsafe to assume, 

 without fuller investigation, that the elements called "cones" 

 are analogous in all Vertebrate eyes. We are dealing here, 

 therefore, only with the structures which are called " cones " 

 in the eyes of the Amphibia, and make no allusion whatever 

 to the structures called '' cones" in the human or other 

 Vertebrate eyes. 



Very superficial examination of the retina of the frog, 

 even when fixed with any of the ordinary reagents, is suffi- 

 cient to show that, judging from the basal limbs alone, there 



