318 PROCEEDINGS OP THE AMERICAN ACADEMY 



not infected with a Rcestelia before the sowings. Considering the 

 abundauce oE Rcestelia on C. tomentosa in this region, such is very 

 likely to have been the case, but, as far as I could tell at the time, the 

 leaves were healthy. For the sake of the argument, let us omit the 

 cultures on C. tomentosa^ arbitrarily assuming that there was previous 

 infection. 



"We have left the fact that G. hiseptatum was followed by spermo- 

 gonia on Amelanchier Canadensis only. Our only Rcestelia growing 

 on the leaves of Amelanchier and not found on other hosts is R. hotry- 

 apifes, found only on the eastern coast from New England southward 

 and this it will bo noticed is also the range of G. hiseptatum* 



Turning to G. fuscum var. glohosum, we fiud that sowings of its spo- 

 ridia on apples, and on CratcBgus oxyacantha and C. Douglasii, were 

 followed by spermogonia, especially on C. oxyacantha, where they were 

 very abundant. The result of the sowings of this species are, then, 

 compared with other cultures, an abundance of spermogonia on C. 

 oxyacantha and an absence of them on Amelanchier. Of the Rcestelice 

 growing near Cambridge, R. aurantiaca occurs on species of Cratcegiis 

 and apples; but I have not found it on Amelanchier, although accord- 

 ing to Peck it occurs on tbat host in New York. The forms included 

 under R. lacerata and R. penicillata are common on Amelanchier 

 near Cambridge, while G. globosum was not followed by spermo- 

 gonia on that host. The G. globosum ranges from Canada to Wis- 

 consin and South Carolina. Rcestelia aurantiaca extends from New 

 England to Arkansas, where it was found on a species of Cra- 

 tcegiis by Prof. F. L. Harvey. The distribution of the Rcestelia and 

 that of the Gymnosporangiinn are about the same. The Rcestelia 

 bears its a^cidia usually on the young fruit and stalks, while the sper- 

 mogonia are borne on the leaves. Plence, in cultures of leaves un- 

 accompanied by young fruit, even if there really is a connection 

 between G. globosum and R. aurantiaca, one would naturally expect 

 to get only spermogonia. Remembering that the ascidia of R. auran- 

 tiaca develop on the berries rather than the leaves, I have tried to 

 obtain the young berries for my cultures ; but I have never yet 

 found any berries formed at the time when the Gymnosjwrangia 

 were ripe. 



* III my paper on Gyvmosporangia this species is also given on Lil>oce.drus in 

 California, on tlie authority of Ilarkness and Moore. I liave never examined 

 Californian specimens whicli, judging from the host, may belong to a distinct 

 species. G. sprciosum Peck, nearly related to G. Ellisii, occurs on Juniperus 

 occidentalis in Colorado. 



