ZOOLOGY AND BOTANY, MICROSCOPY, ETC. 
317 
several plants, especially of Allium Cepa , Hemerocallis fulva, and Roripa 
amphibia. 
In the loose knot stage, a sharply defined hyaline sphere is formed 
round the nucleus, growing independently, and elongating, without any 
connection of this structure with the cytoplasm being perceptible. While 
the nucleus still retains its membrane, the fibres of the central spindle 
develop in the interior of this hyaline structure in a meridional direc- 
tion from the poles ; the nucleole disappearing synchronously with the 
development of these spindle-fibres, and the membrane of the hyaline 
structure also disappearing soon after the chromosomes have begun to 
assume their equatorial position. 
Extranuclear nucleoles exhibit great variability in numbers and 
position during karyokinesis. They are frequently found in close con- 
tact with the newly formed cell-wall, where they have apparently been 
differentiated out of the unused material of the combining fibres. 
The number of chromosomes is constant in the embryonal tissue (in 
Allium Cepa 12), becoming reduced in older tissues (in Allium Cepa 4). 
In pathologically abnormal cells the number is much larger. Fragmen- 
tation of the nucleus was observed in the cells which become differentiated 
into tracheae. 
Fusiform Proteid Substances in the Balsaminese.*— Sig. G. Amadei 
has studied the peculiar proteid substances of a fusiform shape which 
are widely distributed through the genus Impatiens, though apparently 
not extending to all the species. According to the classification of 
Zimmermann, they belong to the class of proteids which occur in the 
cytoplasm or cell-sap. They are always found in the ovary from the 
period of flowering till maturity ; in I. JBalsamina and Sultani they 
occur in the epiderm of the outer wall of the capsule ; in I. parvijlora 
and glanduligera in the swelling tissues, but not in the epiderm. They 
occur further in the groups of cells which accompany the sieve-portion 
of the conducting bundles, but never in the root nor in the lower portion 
of the stem which is destitute of leaves. These structures vary in form 
with the age of the cell, apparently from the action of the acid constitu- 
ent of the cell-sap ; in older tissues the originally crescent-shaped bodies 
frequently become circular, apparently from further curvature and the 
joining of the two ends. From the frequent occurrence of these struc- 
tures in organs containing chlorophyll and in the neighbourhood of the 
sieve-portion of the conducting bundles, and from their sudden disap- 
pearance in the stamens at the time of the ripening of the pollen, the 
author infers that they must play an important part in metabolism. 
Archespore and Nucleus of Bignonia.f — According to Mr. B. M. 
Duggar, the mature archespore of the microsporange occupies a single 
boot-shaped layer. The primitive archespore is differentiated by peri- 
clinal divisions in certain regions of the outer layer of periblem. The 
tapete on the outer side is cut off by the next periclinal division of the 
hvpodermal layer. The definitive archespore is formed by not more 
than a single division in some or all of the primitive archespore cells. 
The archespore of the megasporange apparently develops no primary 
* Bot. Centralbl., lxxiii. (1898) pp. 1-9, 33-42 (2 pis.). 
t Bot. Gazette, xxv. (1898) p. 111. 
