(4°7) 



mode of life and developed into the form which we recognize as 

 rusts, but rather as to whether the simpler types, as we know them, 

 represent primitive or reduced forms. 



The observations made in the course of the investigations 

 of the genus Gymnosporangium appear to support the latter theory. 

 i. e., that the trend of development is toward reduction in the 

 number of spore-forms and toward autoecism. The facts which 

 have led to this interpretation may properly be considered at 

 this time. 



In most of the groups or tribes of Uredinales the four-spored* 

 condition is common and in many the larger number of species 

 belong to that class, but in Gymnosporangium, as has already been 

 pointed out, a three-spored condition is wholly dominant, repeat- 

 ing spores (urediniospores) being lacking in present known species. 

 It is evident, then, that the species of Gymnosporangium stand 

 apart from four-spored forms either as more primitive or as some- 

 what reduced forms. Certain facts gained from morphology and 

 others from life-histories have assisted in forming a judgment on 

 this point. The roestelia-type of aecium, which prevails in the 

 genus, certainly cannot be looked upon as primitive. With its 

 elongated membranous peridium, made up of highly specialized 

 peridial cells, and its colored aeciospores with their evident germ- 

 pores, the roestelia-type of aecium is surely to be looked upon as 

 a later development than the cupulate-type of aecium, such as 

 prevails in the pucciniaceous forms. This being the case the pres- 

 ence of aecia of the ordinary cupulate-type in the genus Gymno- 

 sporangium may be regarded as a safe means of determining which 

 are the more primitive species within the genus, and in this way 

 we may throw some light on the more general problem. 



G. Libocedri (see G. Blasdaleanum in part 2) a species of western 

 America on Heyderia decurrens (better known as Libocedrus de- 

 currens) has been shown by cultures to be related to Aecidium 

 Blasdaleanum, an aecial form of the ordinary cupulate-type, on 

 Amelanchier and Crataegus. This is the only experimental proof 

 which we have of a relationship between a cupulate aecium and a 

 species of Gymnosporangium, but by analogical reasoning it seems 

 most certain that Aecidium Sorbi, another cupulate form, must 



* In using this expression the writer refers to those forms sometimes called 

 ^-forms. The four spore-forms are pycniospores, aeciospores, urediniospores, 



