21(3 Siiwotli-stciniiicd Form of Fo.njlove 



peloria, as in heptandry, thu degree in which the abnormality is uiaiii- 

 fested, as indicated by the size of the peloiic "flower" has some relation 

 to the order of the axis, whether main or lateral, on which it is b(jrne, 

 as is the case with numerous other abnormalities which exhibit grada- 

 tions. A further remarkable peculiarity observed in all the members 

 (40) of one heptaudrous non-peloric family was the thickening of the 

 margins of the sepals, on which were borne bodies having the appearance 

 of rudimentary ovules. Unfortunately no offspring were raised from 

 these particular plants, and as the character never reappeared in any 

 later crop there is nothing to add to the bare record. 



A serious difficulty all through the work was the almost unavoidable 

 contamination of the whole available ground with a large quantity of 

 self-sown seed. In spite of the precautions taken to meet this contin- 

 gency the appearance of an occasional plant here and there gave ground 

 for suspicion that it was a " stray" and sufficed to render the evidence not 

 wholly beyond question. Eventually, hnwever, by breeding exclusively 

 from white-flowered peloric, heptandrous (i.e. triple recessive)' parents 

 whereby the detection of " strays " red-flowered or normal was at least 

 ensured, unimpeachable records were obtained. These results have now 

 fully established the fJict that the smooth-stemmed condition in Dlr/italis 

 purpurea is a well-defined character- and that pure-bred individuals 

 exhibiting this character, but otherwise identical with the completely 

 hairy plant, breed strictly true. We have in fact in the partially glabrous 

 D. purpurea a form as definite as the partially glabrous (half-hoary) 

 Matthiola incana which I have described elsewhere^, to which indeed it 

 appears to be somewhat analogous. For the half-hoary Stock similarly 

 has a glabrous stem though the type jjlant is tomentose throughout. 



' The recessive nature of peloria was established by Keeble, Pellew and Jones (see 

 'The New Fhytoloyist, Vol. ix. lUlO). In my earlier paper (loc. cit.) I showed that 

 heptandry is similarly recessive to the normal, and the same result has been obtained by 

 Tine Tammes and ShuU (see Shull, Zeitschr.f. indiik. Abstnmmungs- n. Vererbiing.ilehrc, 

 Bd. VI. 1912). The white-flowered plants employed were of the pure white form with 

 greenish spots which behaves as a recessive to the coloured type. 



- Although I have consi.stently referred for convenience only to the stem character, it will 

 be understood that this feature is always associated with the smoother leaf surface though 

 the difference in the leaves is less striking to the eye. And here I may add that the extent of 

 the stem which is smooth is not a delinite proportion of the total length but is determined by 

 the amount of vegetative growth which takes place before the reproductive stage approaches. 

 A practical outcome resulting from this variability is that in many herbarium specimens 

 barely sufficient of the stem is preserved to permit of its character being certainly determined. 



' See Reports of the EvoUition Committee to the Royal Society, Rep. 1, 11102, p. S3, 

 For a fuller account later see ./. o/ Gcneties, Vol. v. No. 3, 1916, p. 145. 



