I30 



SECTIONAL ADDRESSES 



Race has developed in space and time. Starting from a small group of 

 people, who originally lived somewhere near the Caspian Cradl eland, this 

 race has gradually spread out as the millennia passed until to-day it is 

 found all round the margins of Eur-Asia, as well as in North Africa. On 

 the west the Mediterranean zone is complete — -as the rim of ' crater 2 ' 

 suggests. On the east the later Alpines have burst through in their 

 journeys to Polynesia and America. But isolated tribes of the Mediter- 

 ranean Race, such as the Ainu and Lolo, still maintain the growth and 

 spread of their forefathers. (This zone is shown, in plan, in Fig. 7.) 

 Within the ' Mediterranean Crater ' we may picture an ' Early Alpine 

 Crater,' which however is not sketched. Below (and concentric with) 

 Crater 2 we may imagine ' Crater i,' which shows a much earlier racial 

 distribution, that of the negritoes. Here the crater did not develop 

 symmetrically — but extended primarily to the south-west (Congo, etc.) 

 or to the south-east (New Guinea, etc.). However, the general conditions 



CRflTERS OF GRO\NTH 



TocJtorese 



Ne grito /^ace 



>. Ken ti/m Speec h 



''4: ^rrier/ncf Speech 

 '. ?1kd,/-^ Race. 



CflSPIff^ CFf/tOLB 



Fig. 15. — ' Craters of Growth ' are Space-Time diagrams illustrating the irregular 

 spread of a culture-type from the point of origin (as suggested in the corner.) 

 Compare the broken line in Fig. 14. 



of growth are the same as in the other craters, and definitely explain the 

 connection of the Congo negritoes with the Tapiro and Aeta off south-east 

 Asia. 



Turning now to the realm of language, we can illustrate their distribu- 

 tions by the same technique. As I have explained earlier, it seems highly 

 probable that the main language-stems originated in south central Asia, 

 just as did the races — but far later. It must again be emphasised that 

 there is not likely to be any permanent connection between an expanding 

 race and an expanding language. Both may originate in the same area — 

 but language moves so rapidly and is so easily transferred that it is highly 

 unlikely that its limits would long agree with those of race. 



We may if we like picture the three ' craters ' on the right of Fig. 15 

 arranged one above the other, Altaic below, Kentum on top. As regards 

 No. 3, it is suggested that the resemblances of the Indian Dravidian 

 languages to the Altaic (of Siberia) is due to their having a common origin 

 somewhere in the region between. Two ' dead ' languages may have 

 a place in the south-west of this growth-pattern (Fig. 15 at 3). These 

 are Sumerian, spoken near Bagdad in 3000 B.C., and Mitanni, spoken in 

 north Syria between 1550 and 1350 B.C. Both of these, according to 

 some authorities, have resemblances to Altaic and Dravidian. They 



