380 THE BODY AT WORK 



it were, left behind. The two are separated at the moment 

 when the paper starts to right or to left. 



Astronomers have long recognized that one of the smaller 

 stars which catches the attention when they are not looking 

 directly at it may be invisible when the gaze is directed to the 

 spot where it ought to be. It was visible when focussed on 

 rods, but it is not visible when focussed on cones. In most 

 birds the retina shows cones alone. To anyone who for the 

 first time enters a dovecote at night the experience is very 

 curious. A candle is for him a sufficiently strong illuminant, 

 but it does not give light enough to enable the pigeons to see. 

 Although evidently alarmed by the noise made by the intruder, 

 they allow themselves to be taken down from their perches 

 without making any attempt to escape. If, startled by 

 the touch of a hand, they take to flight, they fly against 

 the wall. Pigeons are night-blind. The retina of an owl 

 bears chiefly rods, the outer limbs of which are exceptionally 

 long. 



The outer limbs of the rods are coloured reddish purple. 

 This colour is quickly bleached by light. If a frog which has 

 been kept for a short time in the dark be decapitated, its 

 head fixed for ten minutes in a situation in which a window 

 is in front of it, then carried to a photographic dark-room, 

 where an eye is taken out by red light, opened, and the retina 

 removed, a print of the window will be seen upon it. Such an 

 optogram may be fixed by dipping the retina in alum. 



The retina is easily detached from its pigment-layer. If it 

 has been bleached by exposure to light, it regains its " visual 

 purple " when again placed in contact with its pigment. 

 Evidently the visual purple is renewed from the pigment which 

 lies behind (and around) the rods. 



From the cells of the pigment-layer a fringe of streaming 

 processes depends amongst the outer limbs of the rods and 

 cones (Fig. 30). In a dull light the processes hang but a short 

 way down ; in a bright iight they react almost to the outer limit- 

 ing membrane. They supply pigment to the rods, but their 

 relation to cones is not understood. It is clear, however, that 

 the cones, although they are not coloured, are dependent upon 

 the pigment-fringe, since they always remain in contact with 

 it. Their inner limbs elongate in the dark, lifting them to the 



