OF ARTS AND SCIENCES. 317 



a^cidia were not produced on the seedlings on which spermogonia ap- 

 peared. I did my best to keep the young plants alive, but they were 

 all dead by the end of June. Had it been possible to remove the 

 glasses and expose them to the air, they might have done better ; but 

 in exposing them to the air, one should not forget that he is also ex- 

 posing them to the risk of contact with spores from without. On 

 some of my seedlings the spermogonia were very abundant, and it 

 may be urged that, in such cases, the seedlings were destroyed by the 

 violence of the disease itself before the a3cidia could form. The same 

 objection, however, will not apply to the seedlings on which the sper- 

 mogonia were scanty. Yet the latter died, like the former. The 

 spermogonia appeared well marked on some of the leaves, which 

 after some days dropped off, and were followed by fresh crops of 

 spermogonia on other leaves. In the absence of cecidia, can we infer 

 anything from the spermogonia? 



Before trying to answer this question, I must say that I attach very 

 little value to what I have called the third series of cultures, — those 

 in which shoots of Amelanchier and Pyrus arbutifulia were placed 

 in glasses and the sporidia dropped on the leaves, — for the following 

 reasons. The spores were gathered late in the season, after repeated, 

 showers, so that a mixture of the spores of different species could not 

 be avoided with approximate certainty ; and, furthermore, the spermo- 

 gonia were very few in number, did not always develop on the spots 

 where the sporidia were dropped, but on remote parts of the leaves, 

 and, in one case they appeared so soon after the sowing — two days — 

 that it is much more probable that the shoots were already infected 

 with the Ecestelia before the sowing, than that the spermogonia came 

 in any way from the growth of the sporidia. I think it well, then, to 

 omit from present consideration the cases where spermogonia ap- 

 peared on shoots of Amelanchier and Pyrus arbutifoUa in the cultures 

 of 1883. 



As in previous cultures, so in those of 1883, spermogonia appeared 

 on more hosts, and in greater abundance, after sowing the sporidia of 

 G. fiiscum var. (jlohosum than in the case of the other- species. Tlie 

 poorest result came from G. biseptatum, if we except G. Ellisii, in 

 which tliere was no result at all. In cultures previous to 1883, in 

 which leaves and seedlings of Cratcegus totnentosa were used, spermo- 

 gonia appeared on tliat host after sowing G. macropiis, G. fuscum var. 

 globosum^ and G. biseplatum. G. clavipes was not sown on C. to- 

 menfosa, as spore material could not be obtained at the date of the 

 cultures. It may, perhaps, be asked whether the Cratcegus leaves were 



