GNETACEAE 109 



the cone. The tube becomes filled with a fluid, and the 

 pollen captured at the mouth of the tube is thus enabled 

 to reach the nucellus. Professor Pearson has observed 

 that a hemipterous insect, 0do7ttopus sexpunctatus, visits the 

 flowers, and he thinks that it probably assists in their 

 pollination *. There is scarcely any increase in the size 

 of the cone after fertilisation, nor much change in the 

 colour of the scales"^. The seeds ripen quickly, for the 

 principal flowering season is in January, while the cones 

 are ripe in May. The ripe cones gradually break up and 

 the seeds, enclosed in the membranous winged perianth, 

 are scattered by the wind. If provided with sufficient 

 moisture the seed germinates readily within a week or two, 

 producing a strong elongated tap-root which penetrates into 

 the stony ground or into a fissure of the rock, finally 

 branching into secondary lateral roots. The roots of large 

 plants reach to a depth of 10 or 15 feet, being thus 

 enabled to utilize the apparently insignificant but regular 

 supply of moisture, deposited on rocks and gravel by 

 the mist from the sea, which finally percolates into the 

 subsoil. The annual rainfall in these parts of the Namib 

 is on an average less than one inch, and there are years 

 without any appreciable rain. 



The embryo possesses two cotyledons (Plate 20; Figs. 13, 

 14, 15) which soon perish and the first pair of leaves, which 

 follows the cotyledons, remains the only one, the leaves 

 continuing to grow at their base as long as the plant lives. 

 Gradually the leaves become torn into shreds by the wind 

 and their ends are often buried in the accumulating sand, 

 the base of the leaves bulging upwards. The flat or more 

 or less depressed crown of the stem is generally flush with 

 the surface of the ground, but occasionally the soil or 

 stones or rocks, which surrounded the young plant, are 



* Plant-bugs are generally injurious to plants. 



■ The blood-red colour in Hooker's monograph must be due to some erroneous 

 information. 



