HYGROSCOPICITY 169 



10 per cent., which again is much more above than below the 

 average for vegetable materials, then the variation for the 

 entire seed would be only i per cent., since, as an impermeable 

 hairless seed, it would be non-hygroscopic. Or let us take a 

 permeable seed corresponding to the above in all its characters, 

 except that when deprived of its hairs it shows a hygroscopic 

 reaction of 5 per cent. It is obvious that this reaction would 

 be but slightly increased by the hairy covering, since the in- 

 creased hygroscopicity of the hairs would be concerned only 

 with one-tenth of the seed's weight. 



But in thus depreciating in advance the efFect of hairs on 

 a seed's hygroscopicity we have assumed that the covering of 

 hairs displays a more pronounced hygroscopic reaction than 

 is exhibited by an ordinary hygroscopic hairless seed. My 

 observations and experiments, however, go to show that this 

 is not the case. We will take the impermeable seeds of 

 Ipomoea pes-capra^ which have a dense pubescent covering and 

 an average weight of 2|- to 3 grains. As shown in Chapter IX, 

 the weight of the hairs barely amounts to 3 per cent, of the 

 total weight of the seed. To test the influence of the hairs on 

 the hygroscopicity of the seed two samples, containing each 

 about twenty-five seeds, were experimented on, one with the 

 hairs scraped oflF and the other with the hairs remaining. The 

 first gave a hygroscopic reaction of 0*7 per cent., and the second 

 of I 'o per cent. The hairs, in fact, made little or no difference, 

 and the slight change that occurred was probably connected 

 with quite a diiFerent cause, namely, the occurrence of one 

 or two permeable seeds in each sample, an event which my 

 flotation experiments indicate as not uncommon. 



Still more unexpected were the results of my experiments 

 on the hairy seeds of a species of Gossypi'um, probably G. 

 hirsutum, which are described in the same chapter. The 

 average weight of the entire seed is about I'j grain, and of 

 this nearly one-half, or 44 per cent., is made up by the hairs. 

 A sample of scraped seeds and a sample of hairy seeds 

 displayed a similar but slight hygroscopic reaction of rather 



