TRANSACTIONS OF SECTION K. 575 
in the chloroplast, where it may exist in a form similar to that of the nuclear 
material, chromatin. If we connect this with the fact that in many cells in 
which large quantities of reserves are stored up there is, previous to their forma- 
tion, a migration of nuclear substance into the cytoplasm, the hypothesis is at 
once suggested that the pyrenoids and chloroplasts may not be merely cyto- 
plasmic differentiations, but that in their original formation phylogenetically the 
nuclear substance may have had a share, 
The Centrosomes and Centrospheres. 
A vast literature has grown up in connection with the structure and function 
of these bodies because of the special importance which has been attached to them 
as the originators of the process of nuclear division and of the formation of the 
spindle, and because of the important part which it is assumed they play in the 
phenomena of fertilisation. 
Their very general occurrence in animal cells and their prominence in the 
reproductive processes led plant cytologists to predict that they would be found 
to oecur also in plant cells. But their prediction has not been fulfilled. They 
are frequently found among the ‘lhullophytes and Bryophytes, but in the higher 
plants the evidence is steadily accumulating against them, and such structures as 
have been described by Guignard and others are held to be based upon a mis- 
interpretation of the facts observed. 
Where the centrosome exists it consists of a deeply stained granule or group of 
granules surrounded by radiating fibres. In some cases, as in the Basidiomycetes, 
the centrosomes only become definitely visible as minute dots at the poles of the 
spindle, and are not visible until this is completely or nearly completely formed. 
In other cases, as in Dictyota (Mottier), Ascomycetes (Harper), the centio- 
somes with their radiations are clearly visible at two opposite sides of the 
nucleus in the resting stage, and are in close contact with the nuclear membrane. 
In the Ascus, Harper has shown that the centrosome is in close contact, not only 
with the nuclear membrane, but also with the chromatin net, and it seems probable 
that there may be a connection between them. The spindle fibres are formed both 
in Dictyota and in the Ascus in the nuclear cavity before the nuclear wall breaks 
down. In the division of the daughter-nuclei the centrosome which is carried 
over with each daughter-nucleus appears to divide—but this is not certain—to 
give two new centrosomes for the formation of the new spindle figure. 
From some unpublished observations of my own on Polyphagus the centro- 
somes appear at the poles of the nucleus in the same position as in the nuclei of 
the Ascus, and give rise to intranuclear spindles; but I have not so far been 
able to detect that they are continued from one division to another. They seem 
to disappear at each division stage and to be formed anew by a condensation of 
stainable substance in contact with the nuclear membrane, as the mitotic stage 
again approaches, 
In some plants instead of a clearly defined deeply stained controsome there are 
found at the poles of the spindle larger and more diffused cytoplasmic condensa- 
tion called centrospheres. It is not probable that these represent structures 
morphologically different from centrosomes. They are probably of the same 
nature and perform the same functions. 
The centrosome and astrosphere are concerned in some way in the production 
of the spindle figure. Whether they must be regarded as the originators of the 
spindle, or whether they are merely the result of the condensation of forces at 
work in the cell, is a question which at present is not decided. That they are 
not essential to the production of the spindle figure is proved by the fact that they 
do not exist in the higher plants. In these cases the spindle formation is asso- 
ciated with the accumulation of cytoplasm in the form of cytoplasmic caps at the 
poles of the nucleus, from which the spindle fibres penetrate the nuclear 
cavity ; or, especially in the spore mother cells, from a weft of fibres which 
accumulates around the nucleus. These fibres become grouped together to form a 
