﻿Dodge: Effect of host on Gymnosporangium 521 



a most striking case of uniformity in the characters of the aecidial 

 stage when developed on hosts of different genera. O'Gara (13) 

 and Jackson (8) have been able to infect species in six different 

 genera of the apple family with this rust. Malus, Pyrus, Sorbus, 

 Cydonia, Amelanchier and Crataegus are susceptible. The con- 

 stant uniformity of the aecidia on these different hosts, and their 

 "conformity to type" of well authenticated specimens of this 

 rust are their main reliance in assuming that they are working 

 with Gymnosporangium Blasdaleanum on the Libocedriis. 



On the other hand, Pammel (14) has given some data relative 

 to the variations occurring in the aecidiospores and peridial cells 

 when the aecidia of G. macropus are found on different host plants. 

 The conditions under which his experiments were made are not 

 given. He reports the infection of the wild crab and Crataegus 

 mollis with G. macropus, and figures peridial cells and spores from 

 aecidia on each of these hosts and on the common apple. There 

 are noticeable differences in the size, form, and markings of these 

 cells. He says: "In Crataegus punctata as well as C. mollis 

 the peridial cells extend much longer beyond the surface of the 

 leaf, and they are more finely lacerate. In Pyrus lowensis the 

 cups are shorter, darker brown, and the peridial cells do not cling 

 together as in the Crataegi. In the cultiv^ated apple the cups are 

 smaller than [on] either of the other host plants. They project 

 but slightly beyond the surface." It is a common practice, where 

 the aecidial host is known to harbor several species of Gymnospo- 

 rangium, to consider the peridial cells as the most satisfactory 

 means of determining the species of rust, just as the markings 

 on the ascospores of certain species of Discomycetes are thought 

 to be good specific characters. It is well known that the cells 

 at the apex of the peridium differ in length from those at the base, 

 but the shape and markings have been assumed to be fairly 

 constant. 



Long (11) has also recently reported variations in Puccinia 

 Ellisiana when grown on Viola and on Pentstemon. On the former 

 host the aecidia are from one to four times as long, are greater in 

 diameter, are deeper in color, have more peridial segments, and 

 dehisce more tardily, while the aecidiospores are smaller than when 

 the rust is grown on Pentstemon. 



