182 On the Lycoperdon cancellatum. 



that the seeds were all taken from trees which were not quite 

 free from disease ; and that I saw in the last spring some 

 diseased plants, in a case where every precaution, except that 

 of using new pots (which had been my previous custom), 

 had been taken; and therefore, whilst so little is known 

 respecting the habits of plants of this tribe, the preceding 

 facts are not sufficient to support a decision, that the source 

 of the disease might not have been in the seeds themselves. 

 For as the fructification is probably every thing which is 

 seen of this, and many other parasitical fungous plants, the 

 plant may extend in minute filaments through the whole 

 body of the tree which supports it ; and it appears in this 

 view of the subject possible, that these slender filaments 

 may extend into the seeds. The following circumstances, 

 however, militate strongly in opposition to this conclusion. 

 A great number of seedling Pear trees, which were very 

 much diseased, were removed, in the last spring, from my 

 garden to a distant situation, after having had their roots 

 and stems carefully and repeatedly washed, and brushed, so 

 as to remove from them every particle of the mould, in 

 which they had previously grown ; and upon these not a 

 vestige of disease has since appeared. Grafts also, which 

 were formed of parts of diseased trees, have in all cases 

 produced perfectly healthy foliage, even when inserted into 

 the branches of other diseased trees ; which circumstance I 

 think interesting, because it tends to point out a further appa- 

 rent similarity in the habits of this species of fungus, and that 

 which forms the mildew of wheat ; which ceases to vegetate 

 as soon as the straw is severed from its roots, though that 

 remains for some time green and living : whence arises the 



