The Structure of Dinichthys 



177 



was not developed as an isolated element, but in the same year he himself figured the 

 spinal in this form as a part of the intero- lateral (1890.4). Three years later he also noted 

 Sp in Phlyctaenaspis (1893). 



Woodward (1891.2) wrote on this element in Acanthaspis from Spitsbergen, which 

 he, however, did not regard as an Arthrodire but as an Antiarch. In 1902 and 1903, 

 Jaekel described this element as an isolated plate in Coccosteus decipiens Ag. under the 

 name "Ruderorgan." In 1906 and 1907, he spoke of the same plate in some new forms 

 from Wildungen, proposing to name it spinal. Obrutschew (1927) also mentioned Sp 

 in describing a new Arthrodire from Siberia (Angar- 

 ichthys). It is, however, doubtful whether the element 

 he figures is really a spinal. Finally, Heintz; (1929.1, .2) 

 described it in Acanthaspida from Spitsbergen, and 

 Broili (1929, 1930) in Acanthaspida from Germany. 



In all the previously mentioned forms, this ele- 

 ment is actually developed as a more or less long spine. 

 It is attached between AL and AVL, which overlaps 

 its basis from the upper and under side (Acanthaspis, 

 Phlyctaenaspis, Pholidosteus . In Coccosteus the relation 

 is not yet clear). In Dimchthys, on the contrary, the 

 spinal in no way resembles a spine'. It is a small, 

 strongly-bent plate (Text-figure 64; Plate IX, figures 

 23, 24, 25, and 26) which forms the lower, curved part 

 of Newberry's "clavicular." It is represented by four 

 well preserved specimens in the collection of the Ameri- 

 can Museum (one attached to AL), and by two good 

 pieces in the Buffalo Museum. 



Here the present writer wishes to express his gratitude to the Buffalo Museum of 

 Natural Science for its kind permission to describe this plate, and for the fine photographs 

 of it given him by the Museum (Plate IX, figures 23, 24, and 25,). 



At first glance, the long upwardly-directed, spine-like process (pr) catches the eye. 

 This process serves to consolidate the contact with the hinder front part of AL. It fits 

 into a deep, narrow socket on AL and thus joins these two plates very strongly. A cor- 

 responding arrangement is to be found between PDL and PL It is difficult to say 

 whether this spine is always so strongly developed. The writer has only found it in 

 two examples in Buffalo (belonging to the same specimen). None of the specimens in the 

 American Museum show this remarkable spine. Of course, it may be broken, but it may 

 also be possible that the strong development of the SP in the examples in Buffalo is an 

 individual variation only. 



The upper margin (um) of Sp (Text-figure 65) is more or less zigzag in form, where it 

 comes in contact with AL. The well-rounded front margin (Jyn) is sharply bent. The 

 lower margin (Im) is clearly divided in two parts. The lower (Ip) touches IL. The upper 



Text-figure 64. 



AVL, IL and Sf) in contact with each 



other in two different species of 



Pholidosteus sp. Jk. 



(Sketches made by the author in 1927 from 



Jaekel's original material.) 



