47 



are applicable to every stem, and secondly as a derivative be- 

 sides the general signification resulting from its composition 

 can have a peculiar sense too. The number of affixes existing 

 in Greenlandish can be rated at 200. The number which can 

 be attached to the same stem or embodied in one derivative is 

 restricted by no distinct rule, but hardly exceeds, and very 

 rarely reaches 10. 



I have tried TO CALCULATE HOW MANY WORDS IN THIS WAY 

 COULD BE DERIVED FROM A STEM, but did not complete the 

 experiment on account of the appalling increase of the number 

 on each subsequent addition of an affix. I selected the word 

 igdlo a house, and running over the affixes 1 found about 80 

 of them able to be added immediately to this stem, giving 80 

 derivatives. Again at random I took one of these and found it 

 susceptible of 61 immediate additions. I submitted the 61 

 derivatives of second, or stems of third degree to the same 

 experiment and got 70 derivatives of the third degree out of 

 one of them, and so on I got 8 of the 4th, 10 of the 5th, and 

 10 of the 6th degree. But here I stopped and considered what 

 could have been the result, if each time instead of choosing 

 one, I had tried all the others of the same class too, found 

 them equally prolific and finally had summed up all the numbers 

 obtained in this way within the limits of each class: THE FOR- 

 MIDABLE SUM to which I was led deterred me from completing 

 this arithmetical problem, which would have required all the 

 combinations in question to be actually tried and for this pur- 

 pose all written down excepting perhaps the last class. Such 

 a process would be necessary, as several rules have to be ob- 

 served with regard to the order in which a series even of 

 moveable affixes can be appended to a stem, and especially 

 because the applicability of an affix in eaeh case before all 

 depends on its sense. But even if an ample allowance is made 

 for these restrictions, the remaining number will still be so 

 large THAT ONLY THE AMERICAN SO CALLED POLYSYNTHETIC 



