OF THE XJNITED /STATES. 37 



sporidia of the three Gymnosporangia in question, was in consequence of the presence 

 beforehand, in the leaves, of the myceUum of some Koestelia which was made to develop 

 by the moist condition in which it was placed. I am strongly inclined to favor this view, 

 because in many cases I have from the beginning had my suspicions that the leaves of 

 Crataegus tome^itosa used might contain the mycelium of a Eoestelia. 



H. lacerata is so common in the region about Boston at just about the date of the 

 maturity of the Gymnosporangia that it has been with great difficulty, at times, that I 

 have procured leaves of O. tomentosa which appeared even to the naked eye to be free 

 from the fungus. In some cases pots of the young Crataegus used as control plants, 

 showed a growth of spermogonia without any sowing at all, and it was necessary to 

 reject from the cultures all the pots in consequence of the suspicion which was attached 

 to them. In one case, furthermore, spermogonia appeared on a leaf on the fourth day 

 after sowing, a suspiciously early date, unless one supposes that the mycelium was already 

 in the leaf at the time of sowing. Again, why was it that, with abundance of fresh spor- 

 idia of all our common species, in only one instance did spermogonia develop on any 

 other leaves than those of C tomentosa ? The same pains were taken in sowing, the same 

 care was exercised during the continuance of the cultures, yet in spite of that, sperm- 

 ogonia were only produced, one case excepted, on C. tomentosa, the very plant of aU used 

 whose leaves were in some cases doubtful, and produced, too, by three different Gymno- 

 sporangia, none of which is the species supposed by Oersted to be connected with 

 R. lacerata, our common Roestelia on Crataegus. There is only one thing, viz.: the com- 

 parative frequency with which the spermogonia followed sowings of G. glohosum, that 

 prevents my expressing a strong belief that the results of my cultures indicate that the 

 Roesteliae in question were originally in the leaves used, and did not follow as secondary 

 stages of the Gymnosporangia experimented upon. It must be admitted that the accuracy 

 of Oersted's views with regard to the development of the three Danish species is not so 

 generally acknowledged at the present day as it was a few years ago, ancl the note of 

 Eeess on R. penicillata, and Rathay's recent observations, show that even if Oersted is in 

 general correct in supposing that the Roesteliae are genetically connected with the Gym- 

 nosporangia, he has certainly failed to show the connection in the case of given species. 



Much may be said on both sides of the question of the relations between the so-called 

 aecidial and final forms, but in this paper I have only considered the two comparatively 

 small genera Roestelia and Gymnosporangium. Much more work remains to be done in 

 this country. In the first place, more extended and accurate knowledge of the distril^u- 

 tion of our species is to be desired, and many more cultures must be made. A few sug- 

 gestions may be made with regard to the latter point. For the purpose of procuring 

 pure spores of the different Gymnosporangia, a difficult matter if we consider how many 

 of the species are parasitic on J. virginiana, one might gather specimens in March, in the 

 latitude of Cambridge, and allow them to perfect under cover in the house. G. macropus 

 is hardly likely to be connected with R. botryapites, which occurs on Amelanchier, 

 because the young knots are formed in summer before the Roestelia makes its appear- 

 ance. In studying G. Ellisii and G. biseptatum, which occur on the white cedar, and 

 which are hardly known in the East except where found by Mr. Ellis and myself, it 

 would be well to bear in mind that R. botryapites and R. tratisformans are species which 

 have about the same range. It is also an important matter to ascertain the exact date of 



