THE HAMILTON ASSOCIATION. Q9 



High-German similarly, substituting tenues for the original aspirates, 

 had to employ mediae for original tenues and breaths for original 

 medise. — The mistake of assuming that the so-called Low and High 

 German shiftings were contemporaneous has been shown above 

 (p. 91). The account of the second shifting is quite misleading ; the 

 value of the' explanation suggested for the first may better be con- 

 sidered later. 



3. The anthropologists have shown that from the earliest times 

 men of more than one race have spoken Aryan languages, and there 

 is much to favor the view that the divergences we have noticed 

 are due, at least in part, to the acquirement of the language by a 

 people of another race. Canon Taylor ''* notices a significant fact 

 connected with the second shifting. " By race the North-west region 

 of German speech is largely Teutonic, the Eastern Lithuanian and 

 Slavonic, the central region is Celtic, and the Southern is Ligurian. 

 ...... During the last thousand years German speech has been 



slowly winning back its lost provinces, but without displacement of 



population The Low German speech of the conquerors was 



modified when it was acquired by the native tribes. The primitive 

 Low German dialects are only spoken in those Frisian and Dutch 

 districts which are Teutonic in blood as well as speech." Experi- 

 ence shows the correctness of the following : — " We may take it as an 

 axiom that, whenever a new language is acquired by foreigners or by 

 subject races, there will be a class of sounds which will be pronounced 

 with difficulty and will therefore as a rule be evaded or inaccu- 

 rately pronounced. This is especially the case with the soft and aspi- 

 rated mutes." (V. 5.) The Germanic languages seem to have de- 

 parted furthest from the primitive sound system and to have lost 

 more than others of the primitive inflections. This accords well 

 with the theory that the Germanic peoples were originally non- Aryan. 



4. No theory, however, concerning the general Germanic shift- 

 ing can be considered satisfactory, if it fails to take account of simi- 

 lar changes occurring to a limited extent in several cognate langua- 

 ges. 



(a) In old Iranian/, /, k were unchanged only before vowels and 

 and after sibilants ; otherwise they became spirants, as Aves- 

 tan and Old Persian _/ra, Ski. pra, Gr. pro, l^oX.. pio, 'E. for. 

 In Greek they seem to have retained generally their form of 



(i) Origin of the Aryans, Ch. V. 2. 



