LAMARCKIANA SEMIGIGAS. 565 



which the fourth generation flowered in 1923 in the experimental 

 garden in 6 pure specimens. The likeness was already evident in 

 the rosettes, and the 7 plants were planted in a row. By this means 

 their uniformity was made very striking. The spikes are long and 

 thin, with long internodes, narrow bracts, long thin flower buds, 

 and thin cylindrical fruits. All organs are grayish. The flowers are 

 almost erect, and the calyx tube is correspondingly bent. The petals 

 spread only halfway. The tips of the calyx are bent sideways in the 

 plane of symmetry of the buds, pointing outward in respect to the 

 spike. All of the seven specimens reached the same height, which 

 in mid August reached 1 — 1.25 m. They were richly branched. 

 They only differed from the old race in somewhat narrower leaves 

 and bracts, and slightly more compact flower buds, as was to be 

 expected from the influence of their velutina parent. Among the 

 deviating plants of this group two specimens with 15 chromosomes 

 differed only in being weaker, reaching about one-half the height 

 of the others, and in having thicker fruits, which were almost like 

 those of 0. Lamarckiana. 



The three specimens of cana with 16 chromosomes hardly differed 

 from the main type. Two of them had broader leaves, especially 

 in the rosettes, the fruits being cana-\\ke in one of them and La- 

 mar ckiana-Yike in the other. The third plant developed its grayish 

 color relatively late in the spring. Their flowering spikes were as 

 described, with the exception that the anthers of one specimen 

 were almost barren. The two plants with 17 and 19 chromosomes 

 were also evidently cana, one with broader and one with narrower 

 leaves, and both with smaller and stouter flower buds. They were 

 weak plants, which opened their first flowers only in the last days 

 of August. Three specimens of the cana type stayed in the condition 

 of rosettes. 



There was one candicans among the plants. In early youth the 

 race of this name is almost exactly like cana, and so this plant was 

 considered to belong to this type until it developed its flowering 

 spike, which proved it to be a candicans. Differences between this 

 specimen and the race have not been noticed. 



Prototype lata. — There were only two plants belonging to this 

 type, one with 15 and one with 17 chromosomes. Both differed in 

 many points from the pattern as in having stout stems and shorter 

 leaves. They did not reach the flowering period before the fall, and 

 were mainly recognized by the dense grouping of their young leaves, 

 so characteristic for lata. 



