N. Barlow (35 



4. There is doubtful evidence of a second type of .Short-styled 

 plant, which gives a ratio approximating to 1 Long : 1 Mid : 8 Shorts ; 

 whereas the usual ratio with the same individual Mid is fairly near 

 1 Long : 1 Mid : 2 Shorts. Self-fertilized Shorts in Oxalis have pro- 

 duced Shorts only, but in Lythrum one of both the other forms 

 appeared in a family of ten. 



5. A further well-established ratio in a J\] x S cross is 



1 Long : )( Mid : n Short, 



where n is a number between 13 and 34. This indicates the presence 

 of yet a third type of Mid or Short, but no comparative evidence is 

 available. 



The work was undertaken at Mr Bateson's suggestion, and I wish 

 to express m}' thanks for his helj:) and encouragement, and for the 

 facilities which I have enjoyed at the John Innes Horticultural 

 Institution during the last two years. 



REFERENCES. 



(1) Darwin, C. Different Forms of Flowers on Plants of tlie same Species. 



(2) HiLDEBRAND, F. Bot. Zeitiuig, 1864, p. 2 ; 1871, pp. 416 and 432 ; 1887, 



pp. 1, 17, and 33. 

 Monatsherichte d. Akad. d. Wissen. :u Berlin, 1866, p. 371. 

 Lehensverhaltnisse der Oxalisarten. Jena, 1884. 



(3) Bateson and Gregory. Proceedings of the Royal Society, B. 1905, p. 581. 



(4) DE Vries. Species and Varieties, p. 471. 



(5) CoRRENS, Festschrift der medizin-natttrw. Gesellsckaft zur 84. Vers, deutscher 



Naturforscher und Arzte, 1912. 



Journ. of Gen. iii 



