886 



SCIENCE 



[N. S. Vol. XXXVI. No. 939 



non-astringent. If we take a portion of 

 the pulp into the mouth we do not at first 

 notice the puckery effect. In a few mo- 

 ments, however, this develops and becomes 

 more and more pronounced during several 

 minutes. Under the microscope we can 

 actually watch the process. On adding 

 water to some suitably isolated tannin sacs, 

 the contents swell more slowly than before, 

 but ultimately burst out and form a 

 bubble-like mass on the side of the sac. In 

 a few moments a granular veil is seen de- 

 veloping just beyond the surface of the 

 protrusion, gradually increasing in size 

 and moving away further and further. It 

 would take us too far afield to explain why 

 this peculiar behavior, so it must suffice to 

 say that it is due to escaping tannin, which 

 leaves the mass out of which it has escaped 

 unaffected in shape and size. The matter 

 is quite analogous to the washing out of 

 color from cloth: the color goes, the cloth 

 remains. Repeating our experiment with 

 antipyrine, we now find that, while coagu- 

 lation takes place, the amount of shrinkage 

 is less than before, and the action of the 

 reagent on the tannin within the mass is 

 less apparent for the reason that the gran- 

 ulations are smaller. 



If finally we treat a tannin sac from a 

 quite non-astringent fruit in the same way 

 we shall find that it will swell but little or 

 none at all in water, that the alkaloid 

 causes little shrinkage if any, and that the 

 tannin reaction does not take place at all. 

 That the tannin is still there is, however, 

 apparent if we use other reagents, all of 

 which nevertheless act much more slowly 

 than they are known to do when unripe 

 material is tested by them. The conclusion 

 therefore to which I arrive is that the rea- 

 son we do not taste the tannin in the com- 

 pletely ripened fruit is not because it is not 

 there, but because the jelly-like material 

 which occurs in the tannin sacs along with 



the tannin itself becomes coagulated dur- 

 ing the ripening process, so that the tannin 

 may not escape from it except at a very 

 slow rate — too slowly far to be detected 

 even by the delicate membranes of the 

 mouth. To recall what was said earlier in 

 the hour, the tannin is protected by the 

 jelly, so that the alkaloid can not act on it 

 — and this the more efficiently as the coag- 

 ulation is the more complete. Tannin 

 itself, on the other hand, is not a coagulable 

 material. Although a colloid, it does not 

 have the physical properties of a jelly or 

 mucilage. In a word, we have in the 

 tannin masses of the ripe fruit a sort of 

 vegetable leather, which, like ordinary 

 leather, gives up its tannin only very 

 slowly, as shown by long exposure to water. 

 I have tannin sacs of the persimmon which 

 have been lying in water for over two 

 years, but, aside from the loss of tannin, 

 they remain quite unchanged, and will 

 doubtless do so for years to come. 



But what of the relation of all this to 

 the carbon dioxid? We can form some no- 

 tion of the matter if we step aside to en- 

 quire somewhat into the behavior of this 

 gas. The more ordinary name, carbonic 

 acid gas, indicates that it is an acid, and it 

 is therefore of a class of substances which 

 may exert a coagulating (or flocculating) 

 influence upon various colloids. For a 

 single example, carbon dioxide has been 

 found recently to cause the coagulation of 

 the milk or latex of india-rubber trees. In 

 coagulated latex, the india-rubber occurs 

 as minute droplets which remain individ- 

 ual and separate until some coagulating 

 agent has its way, when they run together 

 to form a continuous mass of india-rubber. 

 This is only one example of the effect of 

 carbon dioxid upon substances in the col- 

 loidal state, and by it we are led to suspect 

 that its role in the artificial ripening of 

 dates and persimmons is referable to its 



