Origin and Development of the Composites. 151 
chromosome number in the Compositae shows some interesting 
points. These numbers are recorded in various lists by Gates (57), 
Winge (114), and Ishikawa (66-67), and all the known numbers are 
given in Table XV. Tischler (109) and Winkler (115) have given 
general accounts of the phenomena, but they are not yet available. 
Winge (114) considers the Compositae specially, distinguishing as 
cardinal numbers 9 for the Anthemideae and 8 for the Heliantheae ; 
his classification of the genera seems, however, to be mainly on the 
chromosome number rather than on any generally accepted system, 
e.g., he puts Beilis in the Anthemideae and Calendula in the 
Heliantheae. Rosenberg (94) mentions the Compositae series of 
numbers and shows how they may arise by unequal heterotypic 
divisions, with or without the formation of subsidiary nuclei with 
a small number of chromosomes. 
The known chromosome numbers of the Compositae are given 
in Table XV, with calculated numbers in brackets. Where no 
gametic number is given the species is apogamous and where a 
(v) is appended to the name of an author he has given verbal 
information to the authority quoted below. 
Considering the numbers as given in Table XV, the first case 
is Senecio vulgaris for which Ishikawa (67) records 19. The 
present investigation shows 5 at reduction division, the metaphase 
(Fig. 71) is very clear and in the telophase (Figs. 59, 72) the number 
again appears as 5, with some indication of an idiochromosome 
which shows also in some of the divisions of the chalazal antipodal 
cell (Figs. 65-66). The anomalous nuclear divisions in the basal 
antipodal show other numbers also, about 20 (Fig. 63) or 10 (Fig. 
64). As 5. vulgaris is an aggregate species it is probable that 
Ishikawa examined some other member of the aggregate, in which 
apogamy or hybridisation may have led to complications. The 
figures 56-76 were drawn in 1914, when the writer was unaware of . 
the only two examples of 5 then known as a chromosome number 
in the Compositae, but quite a number of other examples with 5 
have since been described. 
It will be seen from Table XV that the 9, 18, 27, 36, 45 series 
holds good for all the Anthemideae except Achillea Millefolium, in 
which the number is not known with certainty, and Centipeda 
orbicularis, which has 10. The series 8, 16, 32 likewise holds good 
for the Heliantheae with certain exceptions. 
If we seek any phyletic data in the chromosome numbers, the 
5, 10, 15, 20, 30, 60 series is indicated as the primitive one in 
accordance with the phylesis of the family as shown by other data. 
This is the series of the Senecioneae, with the cardinal number 
