5^8 JOUHXA]. OK C(1MI'AH,\II\ !•; NKIUOJ-Ocn. 



As in all natural schemes of classitication, these minor 

 groups overlap each other. To determine the position of 

 doubtful cases, other characteristics have been tabulated. 



To determine the position of a family within the smallest 

 subdivision, the relative size of the prosencephalon has been 

 considered a convenient criterion. 



A series of either gravimetric, volumetric or linear measure- 

 ments would have furnished the desired data. Not being able 

 to use all three, it was necessary to decide which of these three 

 would best serve our purpose. Both volumetric and gravi- 

 metric measurements possess some advantages, but the errors 

 introduced by attempting to separate the various parts of a 

 small brain are so great that all the advantages of the meas- 

 urements are negatived.(') 



There is also an objection to linear measurements. Dur- 

 ing the process of hardening some brains are slightly flat- 

 tened. In one brain this flattening will lengthen one 

 diameter, while in another brain it may shorten it. It is 

 evident then that ratios based upon any one diameter of the 

 prosencephalon would hot serve our purpose. It is also cer- 

 tain that any pressure which causes an increase or decrease 

 in the length of one diameter will also cause a compensating 

 decrease or in increase in some other diameter. Hence, 

 if the ratios of the relative lengths of three diameters at 

 right angles to each can be conveniently combined, there will 

 result a ratio which will be practically free from errors 

 of manipulation. There are several means by which this 

 combination could be effected. I have used what I consider 

 the most convenient combination. I have multiplied the 

 respective ratios of the length, breadth and depth of the pro- 

 sencephalon to the length of the brain together and extracted 

 the cube root of the product. The resultant ratio represents 



I After iiltcinpling to labulali; the rehitive weights of the olfactory lobes nf various 

 birds, A. Bumm remarks: " Abgesehen davon, dass die Abtrennung der Riechhocker 

 uom iibrigen Grosshirn bei den kleinen Vogel nur schwierig and unsicher gelingt, war 

 auch die von mir benutze Wage fiir die hier in Betracht kommenden niinimalen Gewichts- 

 differcnzen nicht enipfindlich genug." ()|i. cit. p. 436. 



