106 C. CLIFFORD DOBELL. 
imperfectly sketched, can all be brought into a more or less 
perfect series with one another. Beginning with Copromonas, 
we have a nucleus in which chromatic and achromatic elements 
are united in the resting nucleus, and do not separate at 
division. A step further we find the nucleus of Hntosiphon, 
which, although apparently identical with that of Copro- 
monas when in the resting state, shows some separation of 
achromatic spindle and chromosomes during division. In the 
Chlamydomonadina also a similar condition is found, but here 
the chromosomes can be clearly counted. A stage further we 
come to the nucleus of Huglena, in which achromatic division- 
centre and chromatic elements are distinct from one another 
even in the resting nucleus. In Hutreptia the chromatin 
seems, in the resting state, to be less closely aggregated 
round the division-centre, and the outline of the nucleus is 
irregular. Chilomonas (see Awerinzew [1]), shows a 
nucleus (consisting of the neucleolo-centrosome and part of 
the chromatin ?) surrounded by chromidia. Such a condition 
may be derived from that seen in Eutreptia by the loss of 
the nuclear membrane. Finally, we reach the nucleus of 
Tetramitus, in which all the chromatin has left the division- 
centre and lies freely in the cytoplasm in the form of 
chromidia. 
That this is a progressive phylogenetic series I do not 
believe. I think it probable that the Copromonas type of 
nucleus is the most primitive, and the Euglena type the 
most highly evolved. From the Euglena type the Tetra- 
mitus type may have arisen by regressive changes. I mean 
that I believe the inter-relationships should be expressed in the 
form of a V with Euglena at the angle and Copromonas 
and Tetramitus at the ends rather than as a straight 
line. These points will have to be taken into account when 
the Flagellata come to be re-classified. 
Let us now leave the nucleus and consider the flagellum 
and its connections. The flagellum has been classified under 
three headings by Prowazek (86), according to its relations 
to the nucleus. These three classes are : | 
