STUDIES IN THE RETINA. 29 



large basal section of Cg may be the result of many discliarges 

 whicli now, on account of the much greater thickness of the 

 whole structure, no longer cause the appearance of nodes. 

 Be this as it may, there can be little doubt but that the forms 

 of cone which we have drawn can safely be regarded as 

 different phases in the construction of the fully developed 

 cone C4 out of matter extruded through the memb. limitans ext. 



This important conclusion is quite in keeping with what 

 is known of the early development of the rods and cones. 

 They first appear as sac-like protrusions of the retina 

 which push back the pigment cells. These sacs are very 

 tightly packed, and may, I find, be more than one deep. 

 This latter fact points to the existence of successive extru- 

 sions. Beyond these simple thin-walled protrusions I am not 

 aware that the further stages in their development have been 

 accurately followed. The rods and cones are thought to be 

 merely specialised prolongations of the " visual cells," the 

 rods being primary and with simple functions, the cones being 

 secondary specialisations for more important functions.^ The 

 new facts contained in this paper will not only fill up this gap 

 in our knowledge of the development of these the most im- 

 portant of the i-etinal elements, but in doing so reverse the 

 judgment as to their relative value, at least in the Amphibia. 



For the moment, such simple sac-like protrusions of the 

 retina beyond the memb. limitans ext., as the admitted start- 

 ing-point in the formation of rods and cones, are all that we 

 require; for we are justified in assuming that this process of 

 development is also likely to be that of any new rod or cone 

 elements which the retina will require in the course of its 

 growth. New elements, therefore, will be formed by the 

 thrusting down, between those already existing, of bag-like 

 protrusions whenever and wherever there is room for them. 

 The shape of the retina makes it clear that growth could not 

 take place only at the rim ; it must be fairly evenly spread 

 over the whole extent of the cup, in order that that shape 

 may be preserved. Now it is obvious that if once a compact 

 ' Max Schultze, ' Archiv mikro. Aiiat.,' iii, 1867, p. 238. 



