OENOTHERA NANELLA, HEALTHY AND DISEASED. 513 



is of a wholly different type, since it is found within the living cells 

 and changes their growth without killing them. Zeylstra pro- 

 visionally placed it in the group of Micrococcus. 



From these data it is probable that healthy 0. nanella might be 

 obtained by giving them less nitrogen and more phosphate of cal- 

 cium. Unfortunately, however, the nitrogen manure acts as the 

 strongest stimulant, under our climate, to induce them to become 

 annual, and for many reasons it is most desirable to have cultures 

 of annual generations. It is, therefore, necessary to determine the 

 amount of nitrogen and phosphate of calcium which will induce a 

 sufficiently large percentage to become annual, but will not essen- 

 tially heighten their liability to become diseased. 



In the summer of 1911, I made some provisional experiments 

 which show that, by this method, there may be produced almost 

 wholly healthy specimens with the normal stature of the dwarfs. 

 In the first place, I found that every part of the stem, every single 

 leaf and flower, may be normal or diseased, in response to external 

 influences. In the young rosettes of rootleaves the first leaves were 

 formerly always twisted; then came long-stalked normal ones and, 

 after these, the really abnormal laeves with broadened and shortened 

 bases, which often killed the terminal bud before it could make a 

 stem. By giving a large amount of phosphate of calcium, and as 

 little nitrogen as possible, every one of the rootleaves could be 

 grown healthy, with a stalk and a narrow wedge-shaped base. The 

 same was the case with the leaves of the stem, and even with the 

 flowers. The number of the abnormal ones could be brought down 

 to a very few, thereby giving the whole plant the appearance of a 

 healthy condition. All transitions between diseased and normal 

 dwarfs were to be seen in these cultures. 



Moreover, 1 have won beautiful healthy dwarfs by means of a 

 cross from which the other parent was eliminated after the rule of 

 the sesquireciprocal crosses^). I pollinated a dwarf of 0. nanella x 

 biennis with the pollen of an ordinary 0. nanella and got a culture 

 of 0. {nanella x biennis) x nanella = 0. nanella which contained 

 a high percentage of healthy plants. They began flowering when 

 only 20 cm. high, the first flower appearing at a height of 10 cm.; 

 whilst 0. Lamarckiana reached 1.50 m. before flowering, the first 

 flower opening about 80 cm. above the soil. All their leaves were 



I) jjUeber doppeltreziproke Bastarde", Opera VI, p. 504. 



33 



