236 CLASSIFICATION OF ALGONQUIAN TRIBES [BPH, ANN. 28 
The following table shows all the consonantic clusters found in the 
writer’s Arapaho notes of 1912: 
Arapaho 
Initial | ra 
conso- Second member of cluster 
nant 
| | 
ee g ie mer: te s 
eh | | | 
| | 
t | | tm t 
b bg bn bs 
$ Beet st sn s 
x | 19 at rm atc 
It has not been feasible to separate genuine and pseudo clusters. 
The x before ¢ and fe is exceedingly weak. The clusters in the 
writer’s Arapaho notes of 1910 were of the same general character 
but contained 0g, 6d, On, and vn also. No clusters begin or end a 
word. 
It will be seen that the clusters differ fundamentally in character 
from those of Piegan, Cheyenne, and Eastern Algonquian. This fact 
points decidedly to the clusters, with certain exceptions, in all of these 
languages as secondary in nature and not original. 
The grammatical analysis is extremely difficult. It is clear that 
many secondary phonetic changes have taken place in the welded 
verbal compound, and so have obscured the stems. However, a 
sufficient number are clear enough to warrant the assertion that the 
general structure of Arapaho agrees essentially with the general 
analysis of Algonquian given by Dr. William Jones. The instru- 
mental particles occur in the correct position. Of these the writer 
has been able to recognize 6 (Fox, etc., m; no m exists in Arapaho), 
Ty Rt, Ws 
The personal pronouns of the independent mode (with certain 
apparent exceptions in the negative verb) are suffixed. Here is a 
very striking difference between Arapaho and normal Algonquian. 
The fact that the terminations are suffixed (not partially prefixed and 
partially suffixed) suggests that in origin they are conjunctive endings 
(compare Micmac), and so far as the writer has been able to find cog- 
nates at all for them (in only a decided minority of instances), it has 
been with the terminations of thismode. Doctor Kroeber, above cited, 
has noted that Cheyenne n- as the prefix of the second person singu- 
lar, independent mode, apparently corresponds with Arapaho -n. 
This the writer considers improbable, as it would be incredible that in 
Arapaho a verbal pronoun that in all other Algonquian languages is 
prefixed, should be suflixed. 
