QUANTITATIVE METHOD AND PRIMORDIA 45 



shady places (in woods) the pigment does not exist and the grey 

 colour is visible. The spikelets are therefore shmy grey ; in 

 other words, silvery-grey. 



Numerous similar examples will be undoubtedly observed as 

 soon as the attention of the describing naturalists is called to 

 this subject, and the descriptions will become more accurate. 



Dissociation by Gradation^ — I call gradation the variation of 

 a given property along a given axis. This subject is ex- 

 pounded in Part VIII. In this paragraph I content myself 

 with one example. 



In Holcus lanatus (fam. Graminece) the sheaths of all the 

 leaves of a fertile stem are clothed with down. Holcus mollis 

 is described as being not generally so downy. In the latter 

 species, according to my observations (numerous specimens in 

 the neighbourhood of Manchester, August-October, 1916), the 

 primordium downy is variable along the fertile stem, which is 

 here the axis of gradation. Comparing the successive leaves 

 from below upwards to the (terminal) panicle, we observe that 

 the sheath of the lowest leaf is downy ^ ; proceeding upwards * 

 we see that the down gradually diminishes on the successive 

 sheaths. The sheath of the upper leaf is always deprived of 

 down (or hardly downy at its base). In this way in Holcus 

 mollis the property downy is distinctly segregated from the 

 property glabrous (absence of down) when the upper leaf and 

 the lower leaf are compared. 



If the sheaths of the successive leaves of one fertile stem of 

 H. mollis are placed in order into a series according to their 

 position along the axis, the variation of the primordium downy 

 is the same as it would be among the hybrids produced by a 

 cross between H. lanatus and a glabrous species of Holcus, 

 supposing that among these hybrids the extremes (parental 

 properties) downy and glabrous were connected by a series of 

 transition forms. ^ (See § 35.) 



According to the above, the difference between both species 

 may be described in the following way * : — 



H. lanatus : sheath of the upper leaf of the fertile stem 

 downy. 



H. mollis : sheath of the upper leaf of the fertile stem 

 glabrous (rarely downy at its base). 



^Compare: "Segregation of the Parental Characters in Seminal Hybrids 

 by Bud- Variation." DARWIN, The Variation of Animals and Plants under 

 Domestication (second edition, 1875), vol. i., p. 425. 



* The down is still visible during the flowering period, although the sheath 

 is already fading. 



' It would be interesting to know whether a Holcus with glabrous sheaths 

 exists, and to cross it with //. lanatus. 



* The description here given is applicable to the specimens observed in the 

 Manchester district. It would be interesting to repeat my observations in 

 other districts or countries. 



