402 SCIENCE PROGRESS 



result of their becoming adapted to wind-pollination, owing to 

 the presumable absence of the insects necessary for cross- 

 fertilisation. 



Within comparatively recent times the catkins of the 

 willows have reverted to pollination by insects ; in order 

 to attain this end the flowers now secrete honey and have 

 become conspicuous by the increased yellow colour of the 

 stamens. It seems not improbable that the extraordinary 

 variability of Saiix may be in some way correlated with this 

 abandonment of wind-pollination in favour of visits by insects, 

 whilst the poplars, which stand nearest to the ancestral type 

 of the two genera Salix and Populus, retain the habit of anemo- 

 phily and display none of the great variability of the willows. 



With regard to the question of a reduction in parts being 

 influenced by the necessity for an economy in materials, it 

 is interesting to note that it is a distinct economy for the 

 plant to have acquired the essentially Angiospermic character of 

 the protective envelopment by the carpel of the ovule ; for, 

 as Arber ^ points out, the chance of pollination is thereby 

 increased, since the insect has only to leave the pollen on 

 one part of the carpel. The same author - indeed considers 

 that entomophily by means of closed carpels will be found 

 to be the real influence which called the Angiosperms into 

 being. 



The foregoing principle, which would seem now to be 

 substantiated for flowering plants, viz. an increased variability 

 and plasticity resulting from a repetition of similar parts, and 

 a subsequent specialisation by rapid reduction to an optimum, 

 can be applied equally well to the animal kingdom. As 

 Smith-Woodward ^ has succinctly observed, " throughout the 

 evolution of the organic world there has been a periodical 

 succession of impulses, each introducing not only a higher 

 grade of life, but also fixing some essential characters that 

 had been variable in the grade immediately below," and again 

 that " the progress of the backboned land-animals during 

 the successive periods of geological time has not been uniform 

 and gradual, but has proceeded in a rhythmic manner." I will 



' 0/>. cit. p. 74. 2 Op. cit. p. 68. 



' Ann. Nat. Hist, xviii. 1906, p. 312 ; see also his Presid. Address, Geology, 

 Brit. Assoc. 1909. 



