30 PEOF. H. H. SWIXXEETOX AXD ME. A. E. TEUEMAX [vol. Ixxiii, 



The table was then rotated through a certain lumiber of degrees, 

 and another diameter was brought into line Avith the direction of 

 travel. This process was re^^eated until readings had been taken 

 along a sutheient number of radial lines to allow of a contoured 

 plan being made. Though this plan proved of great value, it 

 could not be ]-elied upon to record certain details, inasmuch as any 

 portion of the septal svu'face which tended to be an * under- cut ' 

 could not be registered by the vertically moving needle. 



In the preparation of a plan (figs. 11, 11. & 17, pp. 44, 46, & 49) 

 showing the precise position of successive surfaces produced b}' filing, 

 the following procedure was adopted. A portion of an adult whorl, 

 containing at least one septum, was isolated, and the ends were 

 ground flat and polished along a plane parallel to that of the septmn. 

 The specimen was then placed on the turn-table, with one of its ends 

 pressed against the polished side of a metal block, fastened so that 

 the plane of the septum la}- in the direction of tiuvel, and under 

 the point of the needle. The rounded venter of the whorl was then 

 made to pass under the needle, and readings were taken at regular 

 intervals by means of the scale and vernier. This was repeated 

 for the side and the dorsum. The surface of the whorl was then 

 carefully filed away over the area occupied by the suture-line. 

 Within the zone of complicated frilling, in a specimen of the size 

 of Dacfylioceras commune, a depth of only •25 mm. was removed; 

 but in deeper and less complicated regions of the septum this was 

 increased to 'o mm. or even more. At each level a series of read- 

 ings was taken, and the corresponding jJi'ofile of the filed surface 

 was made and introduced into the jDlan. With small ammonites 

 the whole specimen was used, and a wax or plaster mould was 

 made to hold it. 



For the purposes of general systematic work, upon a large 

 quantity of material, the methods just described involve too much 

 labour and, incidentally, the whole of the septum may be lost by 

 tiling. Such methods were necessary in order to test the value of 

 septal sections of one septum for the first time, and all the results 

 recorded in the following pages Avere obtained by them. For most 

 purposes, it will sufiice if the worker takes a portion of a whorl 

 containing not less than eight septa, and files it into a truncated 

 cone or, better still, into a series of steps (PI. II, figs. 3 & 4). 

 The successive septa will be cut at increasing depths, and aWII 

 show most of the salient features seen in a series of sections made 

 from one st^ptum. As a rule, the differences between successive 

 septa Avill not be great enough to vitiate the results serioush- 

 (compare PI. II, fig. 4 with text-fig. 10, j). 43), and at the same 

 time the series can be preserved for reference. 



The camera lucida was used for drawing the sutures and septal 

 sections of small fossils. The whorl was turned to show successive 

 portions the drawings of which Avere subsequently pieced together. 

 Sutures of larger specimens Avere reproduced in the manner 

 suggested by Mr. S. S. Buckman : i namely, by painting the chamber 

 behind the suture and then tracing the latter. 



1 ' Inf. Ool. Arnrn.' Monogr. Paljfiont. Soc. 1894, p. 380. 



