326 PALEONTOLOGY OF ILLINOIS. 



After the feature of Prof. Srnitt's work just noted, the next 

 important is the value he assigns to the zocecium in the forma- 

 tion of families and genera, the zoarial characters being on the 

 other hand treated as perfectly subordinate. His claim is that 

 the cells themselves, and not the mode in which they combine, 

 furnish the true test of relationship, and the essential basis of 

 a natural classification. 



As Busk's three orders of the GYMNOL^EMATA are founded upon 

 structural peculiarities of the cell, they are accepted by both 

 Smitt and Hinck's. When however it comes to the farther di- 

 visions of the order into families and genera, we find that the 

 new system departs very widely from the classifications of Busk 

 and d'Orbigny. As none of their families are represented in 

 Palaeozoic Rocks, it is not necessary, at this time, to follow the 

 system in greater detail. 



Unfortunately, none of the supporters of the new classifica- 

 tion have in any way sought to account for the Palaeozoic 

 types, so here we find a perfectly free and unoccupied field. 

 With respect to these the first question that arises is, which 

 system or classification gives the best results, that in which 

 the zoa,rial characters are assigned a high degree of import- 

 ance, or the other in which the zocecium alone furnishes the 

 test of relationship. If it was necessary to make an absolute 

 choice between the two systems, I would not hesitate very long 

 before accepting the second, for it can not be questioned that 

 differentiations in the cell or actual home of the polypide are 

 more trustworthy structural variations than the form of the 

 zoarium. Still, since the latter in a large measure must be due 

 to deviations in cell structure, it follows that zoarial as well as 

 zooecial peculiarities should be taken into account in the fram- 

 ing of a classification, and while it also follows that in their 

 relative values, the former must necessarily be subordinate to 

 the latter, we should not overlook the individuality of the 

 zoarium as a continuous whole or entity, since we do not fully 

 understand the comparative significance of zoarial and zooecial 

 characters. 



An extended study of Palaeozoic Bryozoa shows me that in 

 classifying these early types much caution must be exercised, 

 and that no character or set of characters has the same classi- 

 ficatory value throughout when applied to a large number of 



