FISHES 167 



all shades of this animal's colour must be darker. Here 

 we have a clear case where natural selection keeps a 

 thing at a certain height. As a general rule, we do not 

 find it to do this, but to tend upwards} Nothing is 

 absolutely good. The eye of the mammal seems good 

 to us, but the bird shows that there are better ones. 

 However, even if selection remains at a certain level 

 in the case of many qualities, it certainly never does 

 that with quantities, as in that case it would have to 

 weed out plus and minus variations. But as these 

 neutralise each other by panmixis, there is no need 

 for the action of natural selection. 



We know now, therefore, that panmixis cannot reduce 

 the size of an organ. How, then, can we explain the 

 rudimentary organs .'* 



This would be easiest to do, it is clear, with the 

 Lamarckian principle. The organ that is no longer 

 necessary, this theory would say, is no longer used. 

 It grows weaker and weaker by the disuse, and is 

 transmitted to offspring in an enfeebled condition. 

 This continues until the organ entirely disappears owing 

 to the steady inheritance of the results of disuse. 



But we saw above that the wings of the blue-throated 

 warbler, though they are not used, are found in each 

 individual of the strength and size that the long migra- 

 tion requires. This circumstance led us to distrust the 

 Lamarckian principle, and ias we shall conclude in the 

 sixth chapter that it is completely untenable, we will not 

 delay with it now, but try to explain the rudimentary 

 organs by other means. 



^ Or, of course, downwards. 



