Dec.23, iQis Parasitism of Cronartium ribicola 647 



(20) both figure two seciospore chains rising from a large basal cell in 

 Puccinia falcaricB and Endophyllum sempervivi, respectively, and there is 

 some evidence that aeciospore chains often arise in like manner in C. 

 ribicola. It is hardly possible, however, to compare the multiple fusion 

 cells to the central placental cells reported by Richards (49). Inasmuch 

 as the elements of the seciospore chains are generally binucleate, the 

 number of polynucleate spores being relatively small, it must be con- 

 cluded that either the extra nuclei so cotamon iii tlie basal cells degen- 

 erate or the complex basal cell gives off more than one binucleate spore 

 chain. 



The constant similarity in the process of congugate division in the 

 basal cells of the ascium, uredinium, and telium suggests a definite stabil- 

 ity of the nuclei of the dikaryon throughout its existence. The presence o f 

 the centrosome in the resting and dividing nuclei adds confirmation to 

 the reports of this structure in rust nuclei as given by other writers. 

 While it is impossible to make a definite statement from actual obser\'a- 

 tion of the process, it seems perfectly evident that the centrosomes at 

 each pole may divide, thus forming two loci for the chromatin groups 

 approaching the poles — cf. Olive's figures 5, a, and 10, Plate 22 (41). 

 The theory that these two groups of chromatin represent two simple 

 chromosomes, the interpretation of the phenomenon apparently accepted 

 by Mme. Moreau, does not appear tenable, unless it is assumed that these 

 chromosomes are compound and that they break up into their com- 

 ponents at metaphase and reunite at telophase. The observations of 

 Holden and Harper (21), Blackman (2), Christman (4), and Olive (41) 

 and the evidence presented in this paper establish the presence of 

 more than two chromosomes or chromosome equivalents, possibly eight 

 being the haploid number. Moreau 's (57) figures indicate that her 

 preparations were apparently not sufficiently differentiated to show all 

 the details in the chromatin masses. 



The polarization of the nuclei, so evident in nearly all of the resting 

 stages, recalls the condition illustrated in Olive's figure 8, Plate 22 {41). 

 The arrangement of the chromatin with a point on the membrane as a 

 locus for the convergence of the strands is not due to fixation; for it is a 

 constant phenomenon in nuclei killed and fixed under different condi- 

 tions and with dift'erent reagents. The centrosome is apparently the 

 center of attraction. Harper {18) has conclusively demonstrated 

 similar phenomena in the Erysipheae. He has also called attention 

 to the fact that the nuclei of these ascomycetes exhibit a definite regu- 

 larity and stability with reference to the arrangement of the chromatin 

 strands and the polarization throughout the cy to logical chamges which 

 he observed. He believes the chromosomes retain their individuality 

 to a marked degree in all the nuclear processes. The observations on 

 the nuclear structure in Cronartium ribicola herein presented suggests that 



