VISION 1049 



must therefore have a hand in the regeneration of the pigment. Ever, 

 the visual purple of a retina from which the epithelium has been de- 

 tached will, after being bleached, be restored if the retina is simply laid 

 again on the epithelial surface. And it does not seem to be the black 

 pigment of the hexagonal cells which is the agent in this restoration, 

 for it takes place in the pigment-free retinae of albino rabbits or rats. 

 Even a retina isolated from the pigmented epithelium, and then 

 bleached, may, to a certain extent, develop new visual purple in the 

 dark. This is even true when it has been kept in the dark in a saturated 

 solution of sodium chloride, and is then, after washing with physio- 

 logical salt solution, bleached by light. Here the regeneration of the 

 pigment cannot be the result of vital processes, but must be due to 

 chemical changes in products formed from the original pigment by the 

 action of light. No such regeneration takes place in a retina which, 

 after having been bleached in situ, is removed without the pigmented 

 epithelium and placed in the dark; and the only probable explanation 

 of the difference is that in this case the photo-chemical substances 

 from which visual purple can be formed have been absorbed into the 

 circulation, and have so escaped. 



The inner segments of the cones of certain animals (birds, reptiles, 

 amphibia, and some fishes) contain globules of various colours, ranging 

 over almost the whole spectrum, and including, besides, the non-spectral 

 colour, purple. The globules are composed chiefly of fat with the 

 pigments (chromophanes, as they have been called) dissolved in it. 

 The function of these globules is unknown. They cannot be concerned 

 in colour vision, or, at least, they cannot be essential to it, for in the 

 human retina they do not exist. 



The yellow pigment of the macula lutea does not belong to the layer 

 of rods and cones ; it only exists in the external molecular layer and the 

 layers in front of it ; in the fovea centralis it is absent. 



Time necessary for Excitation of the Retina by Light Fusion of 

 Stimuli. Whatever the exact nature of retinal excitation may be, 

 it is called forth by exceedingly slight stimuli. A lightning flash, 



although it may last only th of a second, lasts long enough 



' 1,000,000 



to be seen. A beam of light thrown from a rotating mirror on the 



eye stimulates when it only acts for 5 th of a second.- The 



8,000,000 



minimum stimulus in the form of green light corresponds, as we have 

 already seen (p. 784), to a quantity of work equivalent to no more 



than erg that is, about rn gramme-millimetre, or . milli- 

 io 8 io 10 io 7 



gramme-millimetre, which is the work done by - th of a 



10,000,000 



milligramme in falling through a millimetre; and it cannot be doubted 

 that a portion even of this Lilliputian bombardment is wasted as heat. 

 So quickly, too, is the stimulus followed by the response that no latent 

 period has as yet ever been measured. It is certain, however, that 

 there is a latent period, as surely as there is a latent period in the 

 excitation of a naked nerve-trunk, although this also has never been 

 experimentally detected. The analogies, in fact, between a muscular 

 contraction and a retinal excitation are numerous and close. Like the 

 muscle, the retina seems to possess a store of explosive material which 

 the stimulus serves only to fire off. The retina, like the muscle, is 

 exhausted by its activity, and recovers during rest. Like the muscle 

 curve, the curve of retinal excitation rises not abruptly, but with a 

 measurable slowness to its height, and when stimulation is stopped, 



