1905] Ganong,— On Balls of Vegetable Matter 43 
pieces; break one while dry, and it holds. All I have opened have 
been composed of the same substance all through, and of the quantity 
I have picked up in the past 8 or ro years, only two have had a dif- 
ferent substance in them. One had a portion of a chestnut burr, 
and the other a piece of a woody plant with roots on it — the plant 
was dead, however — this one is egg shaped. As the balls begin to 
form they are very loosely put together but become matted from long 
rolling. I find them the latter part of August and first of September, 
only in the one section of the pond, and then only when the pond is 
S As to their keeping qualities when dry, there is one in the 
possession of the Concord Antiquarian Society that Thoreau found, 
and he has been dead 39 years. 
Another account of the formation of the Sandy Pond balls has been 
sent me by Mr. Joseph S. Richardson of Concord, Mass. He con- 
firms Mr. Hosmer’s statements excepting as to the precise place of 
occurrence of the balls, which Mr. Richardson places in a cove at 
the north side of the pond. More recently I have been able to see, 
through the courtesy of Professor B. L. Robinson of the Gray Her- 
barium, a MS. description of these balls, without date, written by 
William A. Hart and now in the library of the Gray Herbarium. 
After describing their formation from Eriocaulon stems, the author 
adds: 
The broken stems first collect in the little ripples of the sand ; then 
a changed direction of waves gives them a new motion and they soon 
form into spheres or cylinders. After attaining some size they roll 
freely about, in water about two feet deep and less, on the bottom 
and continue to collect more material. Naturally they form in coves 
where the eddying of waves is suited for the varied motion necessary 
to form them. They were discovered by the writer in Sandy Pond, 
Lincoln, Mass., and so far as he can learn occur elsewhere only in 
Lake Geneva, Switzerland. 
This reference to the occurrence of such balls in Lake Geneva is 
not confirmed by any information I possess, nor do I know Mr. 
Hart's authority for the statement, which, however, is very likely 
correct. 
Yet another, and in some respects markedly different description 
of the Sandy Pond balls has been sent me by Mr. Lincoln Ware 
Riddle, of Cambridge, Mass., who writes : — 
In the summer of 1898 I found over twenty specimens on the 
northern shore of Sandy Pond, Lincoln, Mass. They are called by 
the people of that neighborhood, “ Burr Balls.” There is compara- 
tively little Eriocaulon growing in Sandy Pond; but the shores, espe- 
cially to the north are bordered by a large number of chestnut trees. 
