632 



THE AMERICAN NATURALIST [Vol. XL 



focusing down on them the remainder of the stiiinj) can be 

 traced to its own nucleus. Hence the rod nucleus invariably 

 retains its early position wholly inside the external limiting 

 membrane. We have been unable to find such ett'ects in the 

 chick as (\ameron figures in the frog in his PI. 51, Figs. 24 and 

 29, where the nucleus forms a distinct projection beyond the 

 limiting membrane. 



Vakiatioxs IX TiiK Rkl\tivk Rati-: of 1)i:vi:l()pmkvt 

 In the <levelopment of the n>tinn the time of appearance of the 

 different elements not absolutely lixcd. An eye at ten days 

 may have reached the ^-.unv .iemvc of <levclopment as another at 

 twelve, or again it may be no further advanced than some that are 

 two days younger. In the same retina it might be expected that the 

 state of development of one layer would bear some definite relation 

 to that of the others. But this is not the case. When the gan- 

 glion-cell layer has reached the stage shown in Fig. (i, the inner 

 nuclear layer may aj)pcar as shown in that figure, or it may be 



laver reaches a veiy advanced stage before anv cvtoplasniic 

 chan.ues in ih.^ outc. n,i< lear l,.ser ha^e Ur^uu to'appcai at all. 

 Ontheothei hand the md- ,,nd eo.ie. nla^ be deNrlnpin^ rapidlv 

 ^^hiK■ the ..ther Kr.ei. ai.' Mill in the rarliei sta::e. The fact 

 has luTii mentioned before that u hen the niajoMtx of the unU 

 are in an adxan.e.l stai^e of deNelopment others are found ix-ide 

 them whose nu<-lei haNc just ceased dividing and u hose ru<liments 

 have not begun to dcvelo|) at all. Hence from the appearance 

 of a part it is imp()ssil)le to predict the stage of development of 

 any other, for there aj)])ears to be no definite developmental ratio 

 that miulit stM-\e as a crit.Mion The drawings in each case are 

 from the in.>rc tvpical representatives, an.l <rive the appearance 



layer; tiiose next the external limiting meml)rane become the ger- 

 minal nuclei. 



