498 MISS G, L. ELLES OJT THE GEAPTOLITE [Aug. 1 898, 



var. B(=Mcholson's fig. 3 a). — This specimen is about ^ inch (8-7 

 mm.) long; the stipes certainly diverge more than JN'icholson's figure 

 indicates. I think that what he regarded as the ' two central cellules ' 

 must be the sicula and the earliest-formed theca, which so closely 

 resembles the sicula in form. 



The sicula is long and narrow for the greater part of its length, 

 but widens somewhat abruptly near its aperture, where it attains a 

 width of -j^ inch ('5 mm.). In length it commonly measures ^ inch 

 (3-2 mm.). 



The earliest theca is seen on the right in this specimen ; it 

 originates from the sicula at a point very near the apex, and 



Fig. 17. — Variation in fo7'7ns ofDidymogTSii^tiis gibberulus, J^ich. 

 Obverse aspect, nat. size. 



A = specimen from Bassenthwaite Sand-beds. 

 B & C = specimens from Eandal Crag. Skiddaw. 



[A & C are in the Keswick ^Nat. Hist. Museum, and B in the Woodwardian 



Museum.] 



then grows downward. In general form it resembles the sicula 

 very closely indeed ; thus this species illustrates well the view held 

 by some authors that the so- called '' sicula ' is nothing but the 

 earliest theca developed from a ' zooid germ.' 



The first theca (adopting the usual nomenclature) is in contact 

 with the sicula for almost its whole length, but they curve away 

 from each other near their apertures, leaving a space between them. 

 Lines of growth may be detected along their length, and especially 

 near their apertures. The connectiug-canal must be situated very 

 high up and near the sicula-apex ; for the theca which is developed 

 second * (th. 1^), that is, the first on the left of the sicula, seems to 

 be nearly as long as the first on the right (th. 1^) : cf. structure 

 of reverse aspect. 



The length of the earliest thecse is about -^-^ inch (2'1 mm.) ; 

 those developed later gradually diminish in length towards the 

 distal extremities of the stipes, which thus come to have the appear- 

 ance of being gently rounded off. 



2. — The specimen figured by Nicholson as 3 5 differs only from the 

 above in that a filamentous thread is seen to proceed from the apex 

 of the sicula ; this, according to Wiman's recent work,t must be 

 regarded as the true virgula. 



* The nomenclature for the tbecse is the same as that adopted in my paper 

 on Fetalograjjtus & Cepkalograptas aheady quoted. 



t Wiman, ' Ueber Diplograptidse, Lapw.,' Bull. Geol. Inst, TJpsala, vol. i 

 no. 2 (1893) p. 97. 



