42 Rhodora [Marcu 
waves, I suspect, do not so much construct as wear down a material 
which has already acquired consistency. They preserve their form 
when dry for an indefinite period. 
Apparently this is Thoreau’s only published reference to the balls. 
But I am informed by Mr. Francis H. Allen, of West Roxbury, that 
-Thoreau’s manuscript Journals now in his possession in preparation 
for publication by Messrs. Houghton, Mifflin & Co., contain two addi- 
tional references to them, and these references, by Mr. Allen’s kind 
aid and permission, are here reproduced : -— 
Feb. 4, 1857. Met Theodore Parker in the cars, who told me that 
he had recently found in Lake Michigan a single ball five inches in 
diameter like those 1 presented to the Natural History Society, 
though he did not observe the eriocaulon. It was late in the season. 
Apr. 5, 1858. In the proceedings of the Natural History Society 
for December, 1856, there were presented by Dr. H. R. Storer ‘a 
globular concretion of grass said to have been formed by the action 
of waves upon the sea shore’. Were not these some obtained by the 
Hoars or Emersons from F. Pond? 
The reference in this passage is to the Proceedings of the Boston 
Society of Natural History, Vol. 6 (published 1857), page 93, where 
the citation given by Thoreau occurs. The specimen itself, how- 
ever, is not now in the Society's Museum, as I am informed by the 
Secretary, Mr. Glover M. Allen. 
The mention of the balls in a work so notable as Wa/den has 
apparently drawn much attention to them locally, and has led those 
of Thoreau's readers who possess scientific tastes to desire further 
information concerning them. It was for this reason that the first of 
these balls I had seen were sent me by Miss Madge Palmer of the 
Packer Collegiate Institute, of Brooklyn. Somewhat later I received 
a letter (Oct. 5, 1901) on the subject from the late Alfred W. Hosmer 
of Concord (who speaks of them as grass balls), in the course of 
which he writes: 
The eastern side of the pond [Flint’s or Sandy Pond] is a shallow 
place, the bottom of a hard white sand, covered with ripple marks. 
In these ripple marks there collects decayed eriocaulon, or duck grass, 
and the prevailing winds in summer being wes?, these bits of decayed 
grass are gently rolled together, and as the ball forms, it begins a 
wider rolling, gathering new bits on its way, until they form balls 
from one to four inches in diameter. If they keep under water while 
in process of formation, they keep spherical, but if they are washed 
up to shore, the action of the waves flattens them. Break one open 
while wet, and it cannot be made to adhere again and will go to 
