255 



[Chase. 



Belatice Prohability. 



Phyllotactic. Qerber. Hydrogen. Surd I. Surd II. 



Aggregate 21145(10)^8 1763(10)i» 378(10)i* 3816.8 1 



Mean 2.078 1.928 1.815 1.138 1 



Residual 1.476 1.349 1.173 1.044 1 



184. Another Comparative Summary. 



If we takie the percentages of deviation, instead of the fractional 



deviations from exact multiples of tlie several divisors, and divide by the 

 number of hydrogen units wliich most nearly represents each of Clarke's 



atomic weights, we obtain data for computing other probabilities and 

 probable errors, which are given below : 



185. Incipient Phyllotaxy. 



The probable errors both in Note 182 and in Note 184, seem to give 

 more indications than are furnished by the relative pi'obabilities, of surd 

 influence upon atomicity. There is room, however, for a reasonable doubt 

 whether those indications are other than accidental, and it would un- 

 doubtedly be desirable, if it were possible, to find some more satisfac- 

 tory test of probabilities which are so near to the boundary line between 

 normal and casual coincidences. In the aggregate probabilities, hydrogen 

 stands between the surd divisors and the phyllotactic divisors. The 

 latter were tested by elements which are denser than hydrogen, and, 

 therefore, have a greater atomic inertia. Is it not likely that the former 

 may find their rightful province in a more ^ethereal region, either in the 

 primitive "subsidence" of nebulous matter or in the undulations which 

 precede subsidence? Cyclical tendencies may naturally become more 

 marked as solidification increases. 



