FEBRUARY 15, 1901. ] 
All these matters are treated in the lucid and 
attractive style characteristic of the author, and 
the work easily takes its place as a worthy com- 
panion to its predecessors. 
The portion of the work that will be read 
with greatest interest is that which deals with 
the vexed question as to whether the centro- 
some is a permanent cell-organ, comparable 
with the nucleus in point of morphological per- 
sistence. The independent discovery by Van 
Beneden and Boveri in 1887, that the centro- 
some in the Ascaris egg is such an organ, since 
extended to many other cases, was at first 
hailed as the most important step taken since 
the establishment of genetic continuity in the 
case of the nucleus. That a like principle ap- 
plies to the centrosome, as was first stated by 
Van Beneden and Boveri, has been widely ac- 
cepted by cytologists, but of late years a marked 
reaction against this view has taken place, a con- 
siderable number of competent observers hav- 
ing been led to conclude that centrosomes may 
form de novo, as well as by the division of pre- 
existing centrosomes. While recognizing that 
the centrosome theory has been in some direc- 
tions pushed too far, Boveri regards the present 
reaction as a backward step. Nevertheless, in 
the course of a highly interesting discussion, he 
makes a large concession to the advocates of 
formation de novo, though his general theory is 
developed with the utmost ingenuity so as to 
save the principle of genetic continuity for which 
he has always contended. He sharply distin- 
guishes between a protoplasmic (cytoplasmic) 
and a nuclear origin of centrosomes. While 
formation of centrosomes de novo in the proto- 
plasm is denied, such an origin is admitted in 
ease of the nucleus, though with qualifications 
that involve only a modification and not the 
abandonment of his original theory. 
The strongest evidence in favor of the cyto- 
plasmic formation of centrosomes de novo is af- 
forded by the observation of the American 
observers, Lillie) Mead and Morgan; but all 
arguments based on this evidence are regarded 
as ‘in high degree vulnerable.’ The multiple 
asters observed in the eggs of Chxtopterus, Ar- 
bacia and other animals are believed by Boveri 
to be ‘almost certainly’ of two kinds, the one 
(polar asters, cleavage-asters) being true asters 
SCIENCE. 
265 
arising through the activity of the egg centro- 
some or its derivatives, the other being ‘ pseu- 
dospheres’ of different nature from the former 
and containing no true centrosomes. All argu- 
ments based on the apparent disappearance and 
reappearance of the centrosomes in the cyto- 
plasm during fertilization (as described for in- 
stance, by Lillie and MacFarland@) are regarded 
as having little weight, in view of the impossi- 
bility of distinguishing the centrosome amid 
other granules when not surrounded by astral 
rays. On the positive side the persistence of the 
centrosome in Ascaris is once more demonstra- 
ted, step by step, throughout the first cleavage, 
and the same phenomenon is for the first time 
fully demonstrated in the egg of the sea urchin 
(Echinus) which, as Boveri emphasizes, is one of 
the most difficult of objects. These cases are 
illustrated by a large number of new and very 
convincing figures. 
When Boveri turns to the nuclear origin of 
centrosomes he takes a different ground, basing 
his conclusions on the absence of centrosomes 
in the higher plants, on the phenomena of di- 
vision in the Protozoa and on the experimental 
evidence brought forward especially by Hert- 
wig and Ziegler. This interesting discusson is 
based primarily on the fact that in Infusoria 
and some other Protozoa a spindle is formed 
from the achromatic substance of the elongated 
nucleus, the bipolarity of the division figure be- 
ing determined by the nucleus without the ap- 
pearance of individualized centrosomes; and 
with this is compared the formation of the polar 
spindle in the eggs of Ascaris. To sucha spin- 
dle the new term ‘netrum’ is applied (v7rpov, 
spindle), and its mode of origin is assumed to 
be a primitive mode of spindle-formation to 
which all other types may be referred. The ac- 
cumulation of substance at the poles of such a 
spindle to form ‘ pole-plates,’ as occurs in some 
Protozoa, represents an incipient centrosome- 
formation, with which is compared the peculiar 
mode of division of the centrosome in Diaulula 
as described by MacFarland. In the latter case 
the centrosome is extra-nuclear, but its mode 
of division closely resembles the phenomena 
observed in Infusoria, the mother-centrosome 
elongating to form bodily the spindle from the 
ends of which are differentiated the daughter- 
