( 4 26) 



cells, it has been found that the species so far recognized divide 

 at first into two classes, one having entirely smooth cells, the other 

 having at least a portion roughened. I have subdivided the 

 latter class upon the nature of the roughness, and have used the 

 terms rugose, verrucose, verruculose, and spinulose to designate 

 the four classes. The terms used to designate the various classes 

 are intended only to be descriptive in a general way. The rugosely 

 sculptured cells are furnished chiefly with ridges or elongated 

 ridge-like papillae in such a way that the effect is that of a surface 

 covered with rugae or folds; the verrucosely marked cells are 

 studded with warty or tubercle-like elevations; the verruculose 

 surfaces are covered with low wart-like protuberances; the spinu- 

 lose cells have diminutive spines or spicules. The markings always 

 cover the entire inner wall extending to the side walls, in some forms 

 reaching clear across, in others, only a part of the distance, leaving 

 the outer part of the side wall and the entire outer wall, with rare 

 exceptions, smooth." 



In the telial stage, the nature of the place where the telia 

 appear is used as a gross character. Some occur on leaves or 

 young branches not producing hypertrophy, others appear on 

 fusiform enlargements of the twigs or branches, while others appear 

 on galls or gall-like excrescences. The form of the telia is also a 

 gross character of prime importance. The color of the mature telia 

 before expansion and germination has been taken account of and 

 is in many instances a good diagnostic character. No matter 

 what the color is originally the telia always become yellowish upon 

 germination. The telia are naked and the microscopic characters 

 therefore lie wholly in the teliospores and pedicels, and consist 

 chiefly of size, number of cells, location and nature of the germ 

 pores, and shape and thickness of the pedicels. It may be said 

 here that it has been pointed out by a number of workers (Kor- 

 nicke, 1877; Dietel, 1889; Richards, 1889; Blackman, 1904) that 

 some species of this genus possess two sorts of teliospores in every 

 sorus, thick-walled ones and lighter colored thin-walled ones. 

 Generally speaking these do not represent two classes but only 

 two extremes, for all gradations can be found between the thickest 

 walls and the thinnest. No attempt has been made to separate 

 these in the specific descriptions and the measurements given are 

 to be interpreted as the general average, not including, as a rule, 

 either extreme. The theory has been advanced several times that 



