646 Journal of Agricultural Research voi. xv, no. 12 



cell does not divide again, being in itself the young teliospore (PI. 59, X). 

 The two nuclei appear to become fully reorganized (PI. 59, Y). They 

 then fuse immediately to produce the single large nucleus of the mature 

 teliospore, during which process the nuclear membranes are absorbed at the 

 points of contact, so that the nuclear contents are free to mix (Pi. 59, Z). 

 The fusing contents round up and become surrounded by the reorganizing 

 membrane. In Plate 59, Z, the two centrosomes and the two nucleoli are 

 still visible. The process of the fusion of the latter was not observed. 

 The fusion nucleus increases in size in a remarkable manner immediately 

 after the union of the two contributing nuclei ; and the enlargement may 

 continue until the volume of the fusion nucleus is as much as four times 

 the combined volumes of the two contributing nuclei. Its diam.eter at 

 the time of maximum enlargement nearly equals the diameter of the 

 teliospore. The chromatin at this stage stains very heavily and appears 

 to be condensed into a heavy irregular spireme-like structure, in which 

 the individual strands are frequently knotted (PI. 59, AA). Holden and 

 Harper {21), in discussing the fusion nucleus in the teliospore of Coleo- 

 sporium sonchi-arvensis Lev. (C solidaginis (Schw.) Thum. ?), present 

 evidence that the heavy skein breaks up into long pieces, which later split 

 longitudinally into liner threadlike units. In Cronartium rihicola there 

 certainly occurs a marked reduction in the thickness of the chromatin 

 threads, but no process which could be interpreted as actual splitting was 

 clearly seen. As the chromatin becomes more fmel}^ drawn it contracts 

 into a more compact tangle, the nucleus shrinking meanwhile (Pi. 59, BB), 

 and finally splits into granules (Pi. 59, CC). The nucleus at this stage is 

 nearly spherical; the nucleolus is rather small, and is generally located 

 near the membrane. The centrosome could not be differentiated in the 

 fusion nucleus with an)^ stain. So far as could be determined by careful 

 examination, it does not reappear until the primary division in the 

 promycelium and in the resting nuclei of the promycelium. 



DISCUSSION OF CYTOLOGICAI. PHENOMENA 



The nuclear phenomena accompan5dng cell fusion and spore produc- 

 tion in Cronartium rihicola clearly confirm the views generall}^ held in 

 regard to rust cytology. Fusion of the gamete cells in pairs in the bark- 

 inhabitating aecium parallels closely similar phenomena in other types of 

 secia and analogous sori, while multicellular fusion is perhaps much more 

 common than in other forms investigated. Fromme {14) has called 

 attention to triple-cell fusions and the fact that cells below the fertile 

 layer often contribute to the multinucleate fusion cells of Melampsora 

 lini. The behavior of the fusion cells in C. rihicola shov/s that cells below 

 those of the fertile layer are potential gametes. It is quite certain that 

 the multiple fusions observed are regular occurrences in gecia of all sizes 

 and shapes, whether on roots or steims. Dittschlag (zi) aind Hoffman 



