Febeuabt 12, 1909] 



SCIENCE 



265 



more (usually 3 to 6) separate fixations from 

 each plant. In every case but one, referred 

 to later, the number remained constant for 

 each individual throughout the several fixa- 

 tions. Corroborative evidence was also 

 secured in a number of the exceptional forms 

 by a study of tapetal cells. Hundreds of 

 figures were carefully examined to determine 

 whether individuality in size and shape of 

 chromosomes could be recognized, but no evi- 

 dence whatever was secured to demonstrate 

 that any such differences exist. 



In the majority of these hybrids the exact 

 number of somatic chromosomes has been da- 

 termined with certainty ; but owing to the fact 

 that a few of the intermediates and several of 

 the gigas-Uke hybrids yet require the study of 

 more sections to settle the question of one or 

 two chromosomes more or less, it will be neces- 

 sary for the present to group them merely 

 under the heads of lata, gigas and inter- 

 mediate. The identifications are based both 

 on chromosomal and vegetative characters. 

 While these groupings are sufficient in the 

 main for all classes, II. and III. can well 

 be further subdivided with respect to external 

 characters. 



Class I., lata, is represented by 2 individuals 

 appearing as true lata in every point, indis- 

 tinguishable from lata mutant from earliest 

 seedling stage, and having lata number of 

 chromosomes — 15. Both annuals. 



Class II., gigas, consists of 6 plants having 

 gigas number of chromosomes (30 in each 

 case so far definitely counted). Two of the 

 6 plants were annuals ; one resembled gigas 

 far more strongly than any other hybrid com- 

 ing to flower in the garden, yet hardly to be 

 classed as a good t3rpical gigas. The second 

 annual resembled the individuals of Class III. 

 in a few, and pure gigas in the majority, of its 

 vegetative characters. A peculiar exceptional 

 circumstance is connected with the micro- 

 scopic study of this plant. Fixations of root- 

 tips, prepared April 3, gave an intermediate 

 number of chromosomes. Another dated 

 April 11 showed the gigas number with equal 

 clearness, while a third was poorly fixed and 

 gave no hint to settle the question. A fourth 

 fixation of tapetal cells was made in August 



and clearly demonstrated the higher number. 

 It therefore seems probable that the first fixa- 

 tion was taken from a member of Class III., 

 and the confusion arose from an error in 

 labeling. It is upon the basis of this con- 

 clusion that I have included this plant within 

 Class H. rather than III. 



The third plant under this head (biennial) 

 was characterized in early seedling and rosette 

 stages by whitish markings, so conspicuous in 

 alhida at corresponding stages of development. 

 It differed markedly from the latter, however, 

 in shape of leaf. The three remaining 

 biennials differ considerably from one another, 

 but equally pronounced differences have been 

 noted among individuals of pure gigas. The 

 members of this class, however, form the least 

 homogeneous group of the three with respect 

 to external characters. 



Class III. is represented by 32 individuals, 

 21 of which were annuals. A portion of these 

 had 22, others 23 and some possibly 21 

 chromosomes, although evidence is not yet 

 complete in regard to the last. With re- 

 spect to vegetative characters, the plants of 

 this group fall readily into 3 subdivisions: 



1. Consists of a single individual, remark- 

 able for its narrow leaved, gigas type of foli- 

 age, utterly unlike that of lata. A single lateral 

 flowering bran'ch produced a few buds and 

 flowers which were noted as " intermediate." 



2. Is composed of those individuals which 

 may be classed as true intermediates and 

 includes 12 of the 21 annuals of Class III. 

 The main features characteristic of this group 

 are as follows : resembled lata in imperfect 

 unfolding of petals, scarcity of pollen pro- 

 duced, sterility of pollen and shape of first 

 buds; resembled gigas in size of corolla (with 

 several exceptions), and various parts of 

 flower, in red tinting of sepals (particularly in 

 latter part of the season), and in general 

 vigor of the plant. 



3. Is in many respects the most interesting 

 of all. It is composed of those individuals of 

 class III. that have long slender buds, flowers 

 with smooth petals and yield a moderate 

 abundance of poUen. It may again be prop- 

 erly subdivided into a and h ; the former con- 

 sisting of three plants attaining the height of 



