480 BOTANICAL GAZETTE [DECEMBER 
that the first division of the ascus nucleus is preceded by a well- 
marked synapsis phase, which the most recent zoological and botanical 
investigations have shown to be the most characteristic and impor- 
tant phase of the heterotypical division. While Marre finds a synap- 
sis similar to that described by STRASBURGER (:04), HARPER and 
GUILLIERMOND have found the phenomena to be essentially the same 
as in the pollen mother cells of the flowering plants studied by GrE- 
GOIRE (:04), BERGHS (:04, :05), ALLEN (:04, :05), ROSENBERG 
(:05), STRASBURGER (:05), Mryakr (:05), OVERTON (:05), TISCH- 
LER, (:06), and Carprrr (:06). Harper finds’ permanent central 
bodies in the nuclei of Phyllactinia, and that the chromosomes are 
permanently attached to the central body and are thus brought 
side by side in nuclear fusion. On this ground he concludes that the 
chromosomes are permanent structures, and that they must be bivalent 
in the nuclei of the young ascus, due to the earlier fusion of the sexual 
nuclei. These bivalent chromosomes, he holds, further unite in syn- 
apsis to form quadrivalent structures, consisting of four somatic 
chromosomes arranged side by side, thus accounting for a numerical 
reduction just as in the higher plants. Marre says that the first 
division of the ascus nucleus is heterotypical, while the second is 
homeotypical, which opinion GuILLIERMOND also holds. HARPER 
gives no opinion as to which are the reduction divisions. He has 
pointed out, however, the universality of the occurrence of the double 
division, following synapsis in the spore mother cells of all higher 
plants, as necessary to accomplish chromosome reduction, where the 
chromosomes are bivalent structures. I might also call attention in 
this connection to the elimination of the double division in embryo sac 
mother cells of parthenogenetic angiosperms, discovered by JUEL (:00, 
705), MuRBECK (:01), OvERTON (:04), and STRASBURGER (:05), 
where reduction is not completed. 
HarPeER also points out that the universal triple division occurring 
in the ascus, no matter how many spores are to be formed, is probably 
to be associated with a quadrivalent character of the chromosomes 
in the ascus nucleus. Where one nuclear fusion occurs, as in most 
plants and animals, a double division is necessary to complete the 
reduction and to distribtite the elements to the daughter nuclei; 
while when two nuclear fusions occur, as in Ascomycetes, a triple 
