J. E. DUERDEN 147 



with a northern hen bearing only the average numbers 36 and 37. No 

 retrogression to these low, grand-parental numbers occurs, so the in- 

 fluence of selection is already manifest, the high numbers tending to 

 dominate the low, and the series shows an approximation to purity as 

 regards high numbers. No doubt can be entertained of the possibility 

 of reaching a still higher degree of purity by continued selective breeding. 

 It is submitted that the results, along with those in the two following 

 series, are sufficiently approximate to be already accepted as evidence 

 that each plume has a separate factorial representation in the germ 

 plasm, and that the variations from 33 to 44 are not environmental. 

 The natural ostrich may well serve as an illustration of the principle 

 embodied in the following remark by Prof H. S. Jennings*: "All 

 thorough work has led directly to this result : that any species or kind 

 of organism is made up of a very great number of diverse stocks, differing 

 from each other in minute particulars, but the diversities inherited from 

 generation to generation." 



However high the degree of germinal purity which may be attained 

 by continued selective breeding in the ostrich it is however problematical 

 whether a uniform number of plumes will ever be procured in the 

 progeny. The slight variations between the right and left wings shown 

 in Tables I to III are an indication that departures of one or two 

 plumes from the expected number will probably always occur. The fact 

 that the plumage of the ostrich is in a degenerative phase, and pre- 

 sumably the genetic factors also, has got to be reckoned with. On many 

 of the above chicks, as also among those of the two series below, one or 

 two diminutive or vestigial feathers are to be found at the elbow end of 

 the row of remiges, showing that retrogression is in process, and some- 

 times a vestigial feather will occur on one wing of a bird and not on the 

 other. As regards these one or two degenerate feathers therefore the 

 genetic factors concerned may be held to be weakening and variable by 

 one means or another, and in this state it is evident that they may or 

 may not gain somatic expression, even on one wing as compared with 

 the other. If this be the correct interpretation it is to be expected that, 

 however purely bred, the same degree of variation will be found among 

 the progeny as between the right and left wings of the individual. If 

 the same zygote can give rise to a marked bilateral variation different 

 zygotes from the same parents may also be expected to produce a like 

 variability in the progeny. 



1 Jennings, H. S., " Observed changes in Heredity Characters in relation to Evolution," 

 Jour. Wash. Acad. Sci. Vol. vii. 1917. 



