294 BUREAU OF AMERICAN ETHNOLOGY [Bull. 170 



natural, compacted, silty deposition levels which occurred over the whole of 

 BN-55 from 1 or 2 feet to 41/2 feet in depth. It is probable that the tops of the 

 cremation pits coincided with one of these deposition layers ; a photograph, talsen 

 by Garth, but not published, of a cut across one of his "floors" shows the same 

 pedologic situation as plate 47, h, of this paper. In the absence of further data 

 they cannot be considered as a cultural trait. 



5. References to the cordage and basketry found are unfortimately lacking in 

 the type of data needed. The rod armor might have been more fully described. 

 The string-twined basketry, a peculiar term in view of the well-developed 

 terminology descriptive of basket weaves, was given to Carolyn Osborne 

 for analysis. Unfortunately, she was not informed that there was a 

 publication contemplated and, having a backlog for analysis, did not 

 examine the piece immediately. It is not a hat, but is the bottom of a Sally 

 Bag (Mason, 1904, p. 439, pi. 168). There is no indication from the Weltfish 

 paper that the basketry which she mentions was of the same type (Weltfish, 

 1932, pp. 113-114). 



It is Osborne's present belief that Garth is on the right track when he 

 connects the cremation pits and shed depositories, and the former had so 

 stated this belief in 1950 (since published, Osborne, 1957). However, the 

 evidence now available, and Garth's is certainly the only excavation evidence, 

 is insufficient to establish the connection. The mud dauber nests are certainly 

 corroborative but as evidence they alone cannot staunch the holes in a leaky 

 methodology which permitted excavation of a unique site without the amassing 

 of the type of objective records generally required. 



6. On page 49 the word "gorget" is used rather than "gorge" which is pre- 

 sumably intended. 



7. Also page 49. The greenish stone, most often used for pounding tools in 

 the area is diorite. Granite is occasionally used. The other pestles are likely 

 andesite. 



8. Also page 49. It should be pointed out that red mercuric oxide is not 

 stable and sublimation destroys the pigment. Consequently it is not possible 

 that the red pigment on the cremated bones could have been other than ferric 

 oxide, unless the bones received post-cremation attention from the Indians. 



9. The tool that Garth calls a basalt hand adz (p. 50, fig. 39, a) which has 

 been found farther up and down the River has been thought by us to be a 

 digging or crude chopping tool. There are no ethnographic data that will help 

 us here. 



10. Garth refers several times to "fiinty rocks" (p. 50, passim). The 

 cryptocrystallines used by the later cultures of the area were largely chalced- 

 onies, agates, jaspers. 



11. There is probably small need to point out the errors on page 51. It is 

 doubtful if many American archeologists would use the old Strong system of 

 projectile-point classification if they were seeking to demonstrate "refined 

 cultural relationships." The Juniper Island site, as Garth mentions, should 

 not be treated as a unit in analysis. None of the material from that site was 

 properly removed. The logic of employing an unknown and dubious quantity 

 as a control, and then placing in the doubtful category any groupings of pro- 

 jectile points that are similar may be syllogistic but it certainly lacks any 

 aspect of scientific analysis. An error in figure 41 has been pointed out (above. 

 No. 2) ; it is probably not necessary to question the validity of comparing burial 

 artifacts and selected material from midden excavations of differing periods. 

 Too, the double listing of 45-WW-5 is confusing. 



Garth lists Osborne (1949) as his source for data on 45-BN-3 and 45-BN-53 

 (incorrectly listed as BN-54). He has a point total of only 26 from these two 



