PERMEABLE AND IMPERMEABLE SEEDS 81 



hygrometric condition of the atmosphere between 99 and 101 

 grains. 



Enough has been said in this connection of the singular 

 ultra-dryness of impermeable seeds. Though illustrated here 

 only by the seeds of Guilandina bonducella^ this capacity is 

 described in the case of several other impermeable seeds in 

 Chapter VI. This quality depends for its maintenance entirely 

 on the impervious character of the outer part of the seed- 

 coverings. There is nothing impermeable in this seed but the Ultra-dry- 

 skin and the portion of the layer immediately beneath it. Let tainedbythe 

 it be pierced only by a pin-prick, and sooner or later the seed S^oV the 



takes up water from the air and ultimately behaves like a seed-coats 

 ifi j j i- i M i i andassoci- 



permeable seed, soon decaying as it lies in the soil, unless the ated with a 



conditions for germination arise. Let, however, the outer 

 covering remain intact, and the embryo within may remain for 

 long periods in a state of suspended vitality. But the imper- kernel, 

 meability of the coverings is not the only conspicuous quality 

 associated with ultra-dryness. We have already seen that in 

 the case of the seed of Guilandina bonducella, although both the 

 coverings and the kernel share this character, it is to the kernel 

 that it more especially belongs. In the sample examined, 

 whilst the water-contents of the seed-coverings amounted to 

 7-6 per cent., in the kernel they were only 4/2 per cent. In 

 Chapter VI it will be shown that this excess of water in the 

 seed-coats as compared with the kernel is a typical character 

 of impermeable seeds, a character that distinguishes them from 

 permeable seeds of the same order, the coats of which as a rule 

 possess a rather smaller water-percentage than the kernel. 



It is to be doubted if any seeds are better fitted to preserve 

 their germinating power for ages when buried in dry soil than 

 the seeds of Guilandina bonducella. Professor Ewart, who 

 employs the synonym of Ctesalpinia bonducella for this seed, 

 places it amongst his group of " macrobiotic " seeds that would 

 last from fifteen to over a hundred years under favourable con- 

 ditions. I fancy that some very old seeds were once tested at 

 Kew, but cannot put my finger on the reference. However, 



6 



