382 DR F. A. BATHER. 



to think that the appearances in D. Sedgwicki are due to a similar cause, and to the 

 fact that in the crushed fossils the region of the theca immediately above this stouter 

 end of the lobe was liable to be broken in and so to form a notch. I conclude that 

 there is no evidence for any actual opening. 



§ 80. The preceding remarks are not meant to imply that there was no hydropore. 

 Search should be made for one on the thecal plates that support the brachial appendage, 

 and particularly between the appendage and the vent. In a single specimen of 

 Dendrocystis scotica one of those plates has near its side margin a rounded projection, 

 which may have been connected with some pore or porous area (§ 136 ; PI. II. fig. 21). 

 [In the Tremadocian species (§ 43) there seems to have been a similar structure. 

 June 1913.] 



§ 81. The Stem may be divided into three regions which merge into one another : 

 proximal, median, and distal. The complete stem was twice the height of the theca 

 or more. 



§ 82. The Distal region is the narrowest, is approximately cylindrical, but tapers 

 gently distalwards with a rapid tapering at the point (PL I. fig. 8 ; PI. II. fig. 19). 

 The columnals are about twice as long as wide, vertically bisected, with the dimeres 

 not exactly apposed but tending to alternate, especially towards the distal and proxi- 

 mal ends of this region. The lumen is much wider in some species (e.g. D. scotica) 

 than in others (e.g. D. rossica). The distal extremity is closed, and presents no sign 

 of attachment (PI. I. figs. 2, 8, 9). 



§ 83. In the Median region the stem widens rapidly, the dimeres are shortened and 

 come to alternate or are pushed apart by intercalated plates, and the lumen increases 

 in relative size (PI. I. figs. 3, 6, 7 ; PI. II. figs. 19, 20). The rate of change and the 

 character of the alternation of the dimeres vary with the species; but in all of them this 

 region essentially serves as a transition from the rather rigid narrow distal region to — 



§ 84. The wide, flexible Proximal region, essentially composed of flanged rings fitting 

 into one another, and enclosing a wide lumen, continuous with the cavity of the theca 

 (PI. I. figs. 3, 7 ; PL II. figs. 12, 19, 20, 25). This is a common plan of stem in the 

 Cystidea, especially in the Heterostelea and the Glyptocystidea, but the interest of 

 Dendrocystis is that it furnishes the successive steps in the evolution of these rings 

 from a number of small plates (see §§ 106, 124, 148). 



§ 85. The stem of Dendrocystis has more than once been adduced as evidence that 

 the Pelmatozoan stem may have originated as a gradual lengthening of the lower part 

 of an attached theca (see, for instance, M. Neumayr, 1889, p. 411 ; and F. A. Bather, 

 1896, p. 297). I now perceive in it a greater tendency to regularity than I had 

 supposed, as well as more distinction from the theca by reason of the thickened rim of 

 the proximal thecal plates. But the former view is not to be hastily rejected — for 

 two reasons. 



§ 86. First, as has just been indicated, the trend of evolution in the proximal region 

 is towards regular annulation and from a number of small irregular plates ; in short, 



