560 Trow . — On Fertilization in the Saprolegnieae. 
nucleus soon after its entry into the oosphere, acquires a distinct centro- 
some and astrosphere. The centrosomes may have been present in the 
antheridia and fertilization-tubes in an unrecognizable condition, but the 
astrospheres are certainly new developments. The astral rays are always 
directed at first to the periphery of the oospore, as shown in Figs. 28 
and 29, and they are associated with a mass of protoplasm having most of 
the characters of an ovocentrum. This mass of protoplasm might be 
regarded as homologous with the receptive spots found in the oogonia of 
the Peronosporeae. Apart from the fact that it appears in Achlya at the 
periphery of the oospore and not of the oogonium, there is good reason 
for believing that its affinity is really with the ovocentrum. One might go 
so far as to call it a spermocentrum to emphasize this point, but for the 
reason that the structure in question is not situated at the centre of the 
sperm. Fig. 27 shows the male centrosome and its astrosphere, with and 
without the associated protoplasm — at a and b. The nucleus is indicated 
separately at c for the sake of clearness. The astrosphere of the male 
nucleus was seen by me first in relatively poor preparations, when it was 
so badly fixed and stained that it was impossible to convince Mr. Pole- 
Evans — a very keen-eyed observer — of its presence. The figures are drawn 
from preparations in which the astrospheres were very conspicuous. These 
male astrospheres, in good preparations, are very striking objects. Mr. 
Pole-Evans made a drawing of one so that I might be able to compare our 
observations. The drawings contained exactly the same number of main 
rays. An inexperienced observer was also asked to examine a section, and 
he noticed the rays of the astrosphere at once. I have not been able to 
discover any record of a similar observation of astrospheres in association 
with male nuclei in plants. 
Typical centrosomes and astrospheres appear to be much more rarely 
developed in plants than in animals. In Zimmermann’s (’96) resume , if we 
exclude, as we must, those cases resting on the evidence of Guignard, 
a very small number of genera are recorded as possessing centrosomes. 
Centrosomes together with their astrospheres have, however, been demon- 
strated very satisfactorily in Sphacelaria by Strasburger and Swingle (’9 7), 
in Fucus by Strasburger (’97) and Farmer and Williams (’96, ’98), in Dictyota 
by Mottier (’98 and ’00) and Williams (’04), in Corallina by Davis (’98) — 
the centrosome being represented in this case by a somewhat indefinite 
centrosphere — in Surirella by Karsten (’00), in Peziza , Ascobolus, and 
Erysiphe by Harper (’95, ’97), in Agaricus by Wager (’94), in Pellia by 
Farmer and Reeves (’94), and in Fossombronia by Farmer (’95). A number 
of genera might be added to these on the basis of older and, yet in many 
cases, apparently thoroughly trustworthy observations. The phenomenon 
appears to be commonest amongst the Thallophytes, especially among 
Fungi, but to be rare in vascular plants, possibly altogether absent. 
