September 29, 1905.] 



SCIENCE. 



401 



very significant in this connection that most 

 of the writers who use ' hammock ' have sjDent 

 much more time in the regions they describe 

 than have those who use ' huznmock ' ; also 

 that some who preferred the latter have ex- 

 pressly stated that the natives always pro- 

 nounced it ' hammock,' and yet their faith in 

 the dictionaries seems to have been too firm 

 to be shaken by this indisputable evidence. 

 In some cases it is almost certain that ' hum- 

 mock ' was put in by the editor or printer, 

 withoiit the sanction of the author,^ though I 

 have indeed noticed one or two cases where 

 the same may be said of ' hammock.' 



As far as my experience in the field goes, 

 the natives in Georgia invariably say ' ham- 

 mock.' I have heard this word in the coun- 

 ties of Chatham, Coffee, Lowndes, Pulaski, 

 Tattnall and Wilcox, and it is doubtless used 

 throughout the intervening ones. If any 

 further evidence were needed, a good map will 

 show a Gulf Hammock (also a post-ofiice of 

 that name) and a Hammock Creek in Florida, 

 and a Hammock Island in Georgia. I have 

 never yet seen ' hummock ' on a map though, 

 nor found any evidence that it is ever used in 

 conversation anywhere (in the sense here indi- 

 cated). As usage fixes the langtiage, it follows 

 that ' hammock ' is the correct form. 



Now as for the definition of this word. It 

 is used for quite a variety of conditions, but 

 from all the evidence obtainable it may be 

 defined broadly as a limited area, with com- 

 paratively dry soil (at least never inundated, 

 and thus distinguished from a swamp), con- 

 taining a large proportion of trees other than 

 pines, and located in a region where ' prairies,' 

 marshes or open pine forests predominate. 

 Topographically a hammock may be either a 

 slight elevation, or a depression, or a slope, 

 and its soil may be sandy, clayey or rocky. 

 The soil is usually rather rich, and the trees 

 growing in it are usually mostly evergreens — 

 though there is probably no one tree which 



^ A case of this kind has occurred in the 

 columns of Science since the above lines Avere 

 Avritten and sent to the editor. In the issue of 

 June 16, in the report of a paper I read before 

 the Torrey Botanical Club in April, I am made 

 to say ' hummocks ' instead of ' hammocks.' 



characterizes all hammocks — and they usually 

 grow so close together as to shade the ground 

 and allow the formation of humus, which is 

 almost wanting in adjacent areas. 



A few varieties of hammocks may be briefiy 

 mentioned. On the coast of South Carolina 

 and Georgia, at least in the vicinity of Savan- 

 nah, a hammock is a low sandy island in a 

 salt marsh, conspicuous for its dense growth 

 of evergreen woody plants; and in the Ever- 

 glades of Florida, according to the accounts 

 of several different explorers, it is a sort of 

 rocky oasis, elevated a few inches above the 

 adjacent prairies, and densely wooded. For 

 these two kinds of places the term ' hummock ' 

 (diminutive of hump) would not be altogether 

 inappropriate, and this fact doubtless accounts 

 for some of the confusion above mentioned. 

 But in central Florida, by all accounts, it 

 seems that a hammock is usually a depression ; 

 while, in the interior of the coastal plain of 

 Georgia it is nearly always a sandy slope form- 

 ing an intermediate zone between the river 

 or creek swamps and the sand-hills which 

 border them. 



The published references to the subject 

 show hammocks to range from North Caro- 

 lina to Florida and Mississippi,' and, like 

 many other interesting things, they seem to 

 be strictly confined to the coastal plain. The 

 natives of other parts of the country seem to 

 have no knowledge of such a word, and as no 

 lexicographers, and few writers of any kind, 

 live in the regions where hammocks occur, it 

 is not surprising that this word should be 

 incorrectly treated in all dictionaries. 



As for the etymology of ' hammock ' (in this 

 geographical sense) I have no suggestions to 

 offer, other than that given by Webster for 

 ' hommock ' and ' hummock.' As a hammock 

 as here defined is always characterized by its 

 vegetation rather than by its topography, it 

 can hardly have anything to do with ' hum- 



"In a paper published by Dr. Arthur Hollick 

 about twenty- fiA^e years ago {Bull. Torr. Bot. Club, 

 7: 14, 18S0) there is a reference to a 'hammock 

 of soapstone and iron ore ' on Staten Island, AAdiich 

 looks like a surprising extension of range ; but Dr. 

 Hollick tells me that ' hummock ' is AAdiat he in- 

 tended to say. 



