THE NORTHWESTERN FLORIDA COAST REVISITED. 517 
Of our season of 1916-1917, some days of November and December, 1916, and 
of January, 1917, we devoted to archæological research among sites near Golden 
Lake, Ark., about thirty-five miles above Memphis, Tenn. The remainder of 
January and part of February, 1917, were consumed in reaching Florida and in 
visiting en route down Mississippi river certain sites that we hoped might prove 
of interest. Notes as to this work on Mississippi river form the concluding part 
of this report. 
The remainder of the season (more than two months) was spent in the inves- 
tigation of Choctawhatchee Bay, Florida, and Choctawhatchee river and tribu- 
tary streams, Florida and Alabama. 
In the autumn of 1917 and the winter and part of the spring of 1918 the expe- 
dition investigated, for the second time, the Apalachicola river and Flint river 
to Bainbridge, Georgia, and continued its work down the Florida coast, from 
5t. Andrews bay to Crystal river, somewhat below Cedar Keys, inclusive. 
In advance of our coming, as usual, search was made by Mr. J. 5. Raybon, 
for many years captain of our steamer, with a companion, along Choctawhatchee 
river from Pate Landing, Alabama, to Choctawhatchee bay, Florida, into which 
the river empties, a distance of 122 measured miles by water, and in addition 27 
miles along Holmes river to the end of navigation on it. Captain Raybon also 
covered for the second time (a search by him having preceded our first visit) 
the northwestern coast of Florida from Pensacola to Cedar Keys. In this way 
knowledge as to additional sites and permission to investigate them were ob- 
tained by us prior to our visit, a great saving in time. 
Choctawhatchee river has its source in southeastern Alabama, pursues a 
southerly course into northwestern Florida, and continuing southerly and south- 
westerly about 94 miles enters the eastern end of Choctawhatchee bay, as stated. 
This stream was investigated by us as far up as Pate Landing, Ala. In addi- 
tion, Holmes river, an affluent, was searched about 20 miles up by us. 
Choctawhatchee river, as it had been reported to us as likely to be, proved 
unproductive from an archeological point of view. The banks, mostly low- 
lying, are wooded, habitations being rarely visible from the water, all of which 
tended to make it extremely difficult to obtain information as to the whereabouts 
of the few small mounds that seem to be the only kind along the stream. 
All mounds visited by us along Choctawhatchee river and on the northwestern 
Florida coast were of sand, unless otherwise described. 
Our thanks are tendered Dr. M. G. Miller, who accompanied the expedition 
as anatomist during the past two seasons, as he has done throughout all our 
archeological work in the South, and has aided in preparing this report and in 
putting it through the press. 
Also we wish to return thanks to Mr. L. B. Smiley, who lent efficient aid as 
general assistant during part of our investigation. 
The warmest thanks of the Academy are extended to the numerous owners of 
mounds and sites along our line of work, who so courteously placed them at its 
disposal. 
