ZOOLOGY AND BOTANY, MICROSCOPY, ETC. 
67 
45 bivalent. The numerous garden species of Japanese chrysanthemums 
are found to have the same number of chromosomes as G. morifolium , 
a fact which appears to indicate that the latter is the parent of all these 
species. Erigeron annuus is parthenogenetic, but E. linifolius and 
E. dubius are normally sexual. In the first species the pollen-grains 
are of unequal size, and each contains a single nucleus. In somatic 
nuclear division 26 chromosomes are present ; the first division of tbe 
embryo-sac-mother-cell is not heterotypic, and the four nuclei share 
equally in the formation of the embryo-sac, which thus has also 26 chro- 
mosomes. Nuclear division in the endosperm shows 52 chromosomes. 
In E. linifolius embryo-sac development is normal, but the antipodals 
are anomalous ; the two cells give rise, by repeated nuclear and cell- 
division, to a row of antipodal cells, each containing one, two, or more 
nuclei. E. dubius has only 9 chromosomes and has an unusual embryo- 
sac development. The first two cell -divisions are normal, but no cell- 
walls are formed, and the four megaspores all share in the formation of 
the embryo-sac, which ultimately contains 16 or more nuclei. S. G. 
Gross Structure of Agar Gel.— C. L. Carey (Bull. Torr. Bot> 
Club, 48, 173-82, 4 figs.). A study of the gross structure of agar gfcls, 
which shows that the lamellated appearance is probably due to rapidity 
of drying and varies with the temperature used. When dried at a 
temperature of 21°-25° C. no such structure is seen, although one or 
two slits appeared. The structure is unaffected by stretching on silk, 
since it was also obtained when pieces were suspended or dried on trays 
in the oven. The finest structures were obtained with 5 p.c. gels ; 
those of lower concentrations dried at 43° C. and 70° C. showed no 
great differences. Drying takes place from the outside inward, and 
therefore the gel has a higher water-concentration inside than outside, 
so that the percentage weight of water and agar at the beginning of tbe 
structure-formation cannot be given accurately. S. G. 
Structure and Development. 
Vegetative. 
Interruption of Endodermis in Dracaena. — A. G. Mann (Proc. 
Roy. Soc. Edin., 1920-1, 41, 50-59, 11 figs., 2 photos). A study of 
Fig. 1. — Steps in the breaking up and displacement of endodermis, e, 
by intrusion of parenchyma cells at x. x 250. 
the secondary thickening in the root of Dracaena fruticosa, with special 
reference to the discontinuity of the endodermis. A series of sections 
F 2 
