loS TAOK SKOCIsnERr, 



According to G. \\'. Mi'ij^rk's view, the tulldwiii^ is tlic correct classification: 



Trilnis 1. Mi/ndocopa. coinprisiiig the tainilifs Ci/pridinidae, Haloci/pridae, and Polycopidae 

 II. Podocopa, „ „ ,, Cypridae, Darwinulidae, Nesideidae, Cytheridae, 



and Cytherellidae. 



C. ClaI"S, on the other hand, divides the Ostracod group directly into families. 

 According to this author the H a I o c y p r i d s form a transitional group between the C y p r i- 

 (l i n i d s and the families grouped under Podocopa. 



W'liieh (if tliese methods of classification is preferable? Is any of them to be 

 regarded as comph>tely right? 



T have tried to answer these questions in the second chapter of the general part of this 

 work, entitled: ..Contributions to our knowledge of the natural system of the O s t r a c o d s." 

 As is seen in p. 101 tlie result of my study iuis been to show that it does not seem quite 

 convenient to adopt any of these three methods of classification quite unaltered. It appeared 

 to me to be necessary to divide G. (). Saus' sectio Mj/orfocopa into two main groups equivalent 

 to Podocopa, Cladocopa and Platycopa. but apart from this Sahs's view lias been accepted. 

 The main classificatinn df the Ostracod group that is employed in tin's work is as follows: 



Sub-order I. Cypridini/orme.'s, comprising all (' y j) r i d i n i d s 

 II. Halocypriformes, ,, ,, Malocyprids 



., .. III. Polycopi formes, „ ,, Polycopids 



IV. Cypriformes, „ ,, C y p r i d s, D a r w i n u 1 i d s, N e s i d e i d s, 



and C y t h e r i d s 

 V. Cytherelliformes, ., ., C y t h c r c 1 I i d s. 



in what order ought these grouj)s to be placed? In other words, is it possible to 

 decide which groups are most primitive? 



As is shown in the above-mentioned chapter in the general part of this work, the 

 facts of the matter are presumably that each group is in a number of respects more primi- 

 tive than the others, in other respects more developed. It is difficult to decide with certaintv 

 which of these groups has the greatest number of primitive characteristics. Under these 

 circumstances it seems to me most convenient to adopt the arrangement u.sed by G. W. MOller. 



