280 THE CANADIAN NATURALIST. [Vol. ix. 



relationship of the Berbers and Guanches, and, in particular, M. 

 E. Pegot Ogier iu the book translated by Frances Locock under 

 .the name of " The Fortunate Isles." It must be confessed, how- 

 ever, that this writer, while asserting " that the Guanches may 

 be put down as exclusively of Celtic origin," does not proceed 

 to the proof of the statement, except by comparing a Guanche 

 temple with similar Celtic remains at Carnac in Brittany. Mega- 

 lithic structures of the same character have been found through- 

 out the Berber area, such as that at Bless in Tunis, described by 

 Frederick Catherwood in the Transactions of the American Eth- 

 nological Society. Jackson, in his account of travels in Barbary, 

 gives special prominence to the Berber tribes who call themselves 

 Zimuhr and Amor, whom he regards as descendants of Canaani- 

 tic Zemarites and Amorites. Of the former he says : " They 

 are a fine race of men, well grown and good figures ; they have 

 a noble presence and their physiognomy resembles the Roman." 

 Writing of the Amor, whom, on account of their bravery, the 

 Sultan Muhamed called the English of Barbary, he says : " When 

 the Sultan Muhamed began a campaign, he never entered the 

 field without the warlike Ait Amor, who marched in the rear of 

 the army ; these people received no pay, but were satisfied with 

 what plunder they could get after a battle ; and accordingly, this 

 principle stimulating them, they were always foremost in any 

 contest, dispute or battle." The names Zimuhr and Amor, 

 together with Gomera, that of one of the Canary Islands, tell 

 strongly in favour of a Sumerian or Cymric connection of the 

 Berbers. Sir Henry Rawlinson, in his Essay on the Alarodians 

 ■of Herodotus, gives the name Burbur to the Accadians (? Sume- 

 rians), and, although the correctness of this is disputed by Pro- 

 fessor Sayce, I am disposed to think that the veteran Assyriologist 

 is rioht. It is at least a remarkable coincidence that links 

 Sumerian Burbur and Zimuhr Berbers by a double nomenclature 

 and without any intention on the part of Sir Henry Rawlinson 

 so to unite the widely separated peoples. The grammar of the 

 Berber has been studied by Mr. Newman and others, and has 

 been denominated sub Semitic, but anyone acquainted with the 

 Celtic tongues knows that they also might be called sub-Semitic 

 in character. The marking of inflexion by internal vowel 

 changes, the paucity of tenses in the verb, and the postposition 

 to the verb of the personal pronoun, are Semitic and not Indo- 

 European. Now the two tenses of the Berber verb, the deriva 



