266 
centrosomes. This case gives the transition to 
the usual types (Ascaris, cleavage, etc.) where 
the extra-nuclear centrosome divides bodily. 
Boveri is thus led to regard the centrosome 
of the higher types as equivalent to the intra- 
nuclear material from which the ‘netrum’ of 
the lower types (Infusoria) arises, and he char- 
acterizes the nucleus of the latter as a ‘cen- 
tronucleus’—a view which is nearly identical 
with that of Richard Hertwig. In the higher 
types in general, individualized and permanent 
centrosomes have been differentiated, and as 
it were emancipated, from the primitive ‘cen- 
ronucleus,’ and as arule lie outside the nucleus 
in the cytoplasm; but the nucleus has in some 
cases retained the power to give rise, upon oc- 
casion, to a karyokinetic spindle (as occurs in 
the polar spindles of Ascaris), or even to pro- 
duce new centrosomes. This, Boveri believes, 
is the case with the egg-nucleus in echinoderms, 
and he would thus explain the division figures 
formed as a result of chemical stimulus. In 
normal fertilization, on the other hand, the 
centrosome-producing power of the egg-nucleus 
remains latent, since the spermatozoon imports 
an active individualized centrosome. How far 
the nuclei of higher forms in general haye re 
mained ‘centronuclei’ and still possess the 
power of forming centrosomes, how far they 
have lost this power, remains to be determined ; 
but Boveri does not consider it probable that 
such a mode of centrosome-formation is wide 
spread. It is worth pointing out that Boveri 
regards as not improbable the view of Calkins 
that phylogenetically the primitive form of nu- 
cleus (centronucleus) may have arisen through 
the union in one body of a cytocentrum and 
other elements (chromatin) originally scattered 
through the general cell substance. 
It is evident from the foregoing that the 
original centrosome theory of Van Beneden 
and Boveri, as commonly understood, has thus 
undergone a considerable modification, which 
will very likely be regarded by some readers as 
a virtual abandonment of that theory. Suchis 
not, however, Boveri’s own view. ‘‘Strictly 
speaking the cases in question do not involve a 
new formation. For even though the centro- 
some may not preexist as an individualized 
structure, it does not arise as something really 
SCIENCE. 
(N.S. Vou. XIII. No. 320. 
new * * * but only by the final transforma- 
tion of a preexisting cytocentrum’’ (p. 193).* 
If, however, we accept the widely held view 
that the achromatic nuclear substance is closely 
related to the cytoplasm, the step does not 
seem very great from the formation of ‘ indi- 
vidualized’ centrosomes de novo out of the 
achromatic nuclear material to such a formation 
in the cytoplasm. Boveri’s denial of such 
cytoplasmic centrosome-formation rests upon a 
series of assumptions, some of which are op- 
posed by the recent discovery that Morgan’s 
‘artificial astrospheres’ may multiply by di- 
vision, even in enucleated egg-fragments. He 
has nevertheless entrenched the centrosome 
theory in a strong position, from which it can 
only be dislodged by a stronger attack than has 
yet been made upon it. 
Other valuable discussions deal with the rela- 
tion of centriole and centrosome, and of centro- 
some and sphere, and with the physiological 
activities and cyclical changes of the centro- 
some. Without attempting to review these in 
extenso it may be pointed out that Boveri holds 
fast to the view that the centrosome passes 
through a regular cycle of changes, during one 
part of which it is a body of considerable size 
within which lies a smaller ‘centriole.’ He 
repudiates some of the interpretations that have 
been placed by other writers upon his own 
earlier observations, and identifies his ‘ centro- 
some’ with Van Beneden’s ‘ corpuscule central,’ 
and not with the latter author’s ‘medullary 
zone’ of the sphere. He overturns Kostan- 
ecki’s and Siedlecki’s contention that the size 
of the centrosome depends merely on the de- 
gree of extraction of the hematoxylin or other 
dye, by the highly important observation that 
when at its greatest size the cemtrosome in 
Ascaris is clearly visible not only in unstained 
material but also in the living object. He be 
lieves that the so-called pluricorpuscular cen- 
trosome, such as he himself and others earlier 
described in echinoderms, is an artifact ; but 
his observations justify some severe strictures 
that are passed on the scepticism of such writers 
as Fischer, who have more than hinted that the 
centrosome itself isan artifact. Fischer’s valu- 
* For a related though not identical interpretation 
see Wilson, ‘ The Cell,’ pp. 111, 215. 
