494 



numerous rays not necessary for vision, are as it were eliminated 

 from the operation, the eye being blind to them, and affected only 

 by such as are reflected backwards to the pupil along the axes of the 

 crystalline columns. 



2. The Crystalline Columns of the Compound Eye. — As stated 

 in my lecture on the retina formerly alluded to,' I conceive the crys- 

 talline columns in the eye of the insect or crab, to act in the same 

 manner as the retinal rods in the spheroidal or simple eye. That 

 they do so, may be held as established by the researches of J. Miiller 

 on the laws of vision in the compound eye. Miiller even refers to 

 the columnar structure of the retina, as presenting a certain simi- 

 larity to the structure or arrangement of the compound eye. F. 

 Leydig, in an elaborate memoir published in Miiller's Archiv in 

 1855, on the structure generally of the Arthropoda, examines mi- 

 nutely the structure of the simple and compound eyes, and ar- 

 rives at the conclusion that the crystalline columns of their com- 

 pound eyes, as well as the corresponding structures in their so- 

 called simple eyes or ocelli, are of the same nature as the so- 

 called rods and cones, that is, the photaesthetic bodies which I 

 have already described in the retina of the proper simple or ver- 

 tebrate eye. But Leydig entirely loses sight of a fact, which if 

 unexplained, vitiates his conclusion as to the physiological iden- 

 tity of the bodies in question. In the annulose or molluscous eye, 

 whether in its so-called simple or compound form, the crystalline 

 columns are directed like the tubes of so many telescopes towards 

 the object, the corresponding nervous filaments passing to them from 

 behind ; whereas the crystalline rods of the vertebrate retina are 

 directed away from the object, that is, towards the back of the eye 

 — ai'e in contact in fact with the choroid, while their nervous fila- 

 ments are connected to them in front, that is, between them and the 

 object. 



On the other hand, il' I am correct in holding that the vertebrate 

 eye is acted upon by those rays only which are reflected from its 

 choroidal surface, I have not only explained physiologically why its 

 retinal columns are reversed ; but I am legitimately entitled, as 

 Leydig is not, to consider them as the homologues of the crystalline 

 columns of the annulose and molluscous eye. 



But the teleological explanation of the opposite arrangement of 

 the corresponding structures in the vertebrate and invertebrate eye, 



