MacMillan: OCCURRENCE OF SPHAGNUM ATOLLS. 13 



2. The vegetation of the atolls differs from that of the pond 

 outside and the inner lagoon. It varies with the development 

 and desiccation of the atoll. 



3. The origin of the sphagnum atolls in the cases studied 

 ' may be ascribed to a season of gradual recession of the waters 



of the pond, followed by a season of comparatively rapid in- 

 crease in area and level. 



4. The atolls first appeared as annular floating bogs separa- 

 ted from the shoreward turf as a result of the original zonal 

 distribution of littoral plants and the rise of the waters together 

 with the favorable concurrence of a group of special and 

 necessary conditions. 



5. Some of the apparent conditions of atoll-formation are 

 (a) a definite maximum size and depth of the parent pond; {b) 

 considerable height and regularity of the banks of the parent 

 pond; (c) a regular and gentle slope of the pond bottom from 

 shore to center; (d) a definite original character of littoral veg- 

 etation when the pond was at low level; (e) a reduction within 

 minimum limits of the lateral pressure and tension of winter 

 ice; (/) a comparatively prompt anchoring of the atoll upon the 

 bottom. 



A number of special problems present themselves at the close 

 of the general investigation, but in this connection it will be 

 unnecessary to enter into them in detail. For example, there 

 might be noted the conditions which determine the average 

 width of the atoll, its distance from shore, the ratio between the 

 size of the intra- insular lagoon and of the pond as a whole, the 

 particular causes of the appearance of given species of plants 

 upon a given atoll, the relation between the vegetation of the 

 atoll and of the pond and between the pond and the lagoon and 

 a number of other matters, the full consideration of which would 

 prolong the discussion beyond bounds. The purpose of this 

 paper has been indicative rather than exhaustive, and having 

 described an unusual plant-formation with such hypotheses of 

 its origin as the facts would seem to warrant, it may properly 

 close. In conclusion it may be not immaterial to note the in- 

 teresting condition that would prevail should such circular 

 peat islands become fossilized subsequent to the complete ob- 

 literation of the ponds by invading vegetation. I am not aware 

 that such circular depositions of peat have been encountered, 

 but apparently their formation is not impossible and I should 

 presume that under favorable conditions they might be demon- 

 strated by discovery. 



