552 Gates. — The Trisomic Mutations of Oenothera . 
1912, and has arisen once from Lamarchiana , four times from lata x 
Lamarckiana , and once from pallescens. When selfed it yields liquida and 
Lamarckiana in nearly equal numbers, and also the mutants scintillans, 
pallescens, cana , and oblonga. It thus appears that while these four simple 
trisomic forms arise more frequently from lata x Lamarckiana than from 
Lamarckiana itself, and frequently produce other trisomic forms as mutants 
in their offspring, yet they never give rise to lata itself. This fact appears 
to be significant as indicating that they are perhaps in some sense modified 
from lata , which appears as the basal form in the series. Thus the records 
show that cana can give rise to pallescens and pallescens to cana , also that 
the same interchangeable condition exists between pallescens and liquida . 
But while any of these eight trisomic forms can apparently arise from lata x 
Lamarckiana, none of them except pallescens apparently can give rise in 
turn to lata. 
That scintillans can also give rise to lata, however, is shown by the 
following record (de Vries, 1913 , p. 248). Eight lata mutants were derived 
from scintillans in a strain of Oe. Lamarckiana obtained from Messrs. 
Vilmorin, Andrieux & Cie, of Paris. Lata also appears (p. 257) in the off¬ 
spring of scintillans in de Vries’s strain of Lamarckiana. In addition, scin¬ 
tillans is said to have appeared 14 times as a mutant up to 1900, and to 
arise equally from Lamarckiana 01* lata. Hence this reciprocal relation 
exists between these two trisomic forms. Further, lata x scintillans ( 1 . c., 
p. 257) produces only lata and Lamarckiana (736 plants) and oblonga x 
scintillans gave 61 oblonga, 13 Lamarckiana, and one lata. In another 
record (p. 265) oblonga x scintillans gave 81 per cent, oblonga, while 
scintillans x oblonga gave 18 per cent, oblonga. It has long been known 
(de Vries, 1909 , i. 390) that a special relation exists between oblonga and 
scintillans, the latter producing a progeny which fluctuates widely but 
usually includes a high percentage (2-15 per cent.) of oblonga in addition to 
scintillans, Lamarckiana, and about one per cent, of lata and nanella. On 
the other hand, mut. oblonga when selfed is found to breed essentially true 
except that it gives occasional mutants, including albida, elliptica , and rubri- 
nervis. It does not apparently produce scintillans, but it has arisen as 
a mutant both from lata and from scintillans x nanella. 
By selfing certain lata plants, Miss Lutz ( 1917 ) has succeeded in getting 
360 seeds from three individuals. Of these seeds 129 germinated and 126 
developed. Three of the lata offspring were again selfed and yielded 
259 seeds, of which 99 germinated. Of the 226 plants so produced 109 
were lata , 8 lata (?), 57 Lamarckiana, 4 Lamarckiana (?), and 23 are said to 
have belonged to seven different 15-chromosome types (names unfortunately 
not given), 4 (belonging to three types) were j 4-chromosome mutants, and 
one each had respectively 16, 21, and 22 chromosomes. The remaining 
j8 plants were apparently not classified. The offspring may be said to 
