120 



Bashford Dean Memorial Volume 



Text-figure 6. 



Branson's reconstruction of Dinichthys terrelh 



Nwb. (1908). 



ADL, antero-dorso-lateral; AVL, antero-%'entro-lateral; CI, 



clavicle; M, marginal; MD, median-dorsal; PDL, postero-dorso- 



lateral; PVL, postero-ventro-lateral; SO, sub-orbital. 



the exoccipital and marginal. In the pos- 

 terior corner of the marginal he figured a 

 new sensory canal, nearly perpendicular 

 to the marginal canal. As we shall see 

 later, Branson has here mistaken the limit 

 between two plates for a canal. The sub' 

 orbital is very large since his sub-orbital 

 was in reality three plates. He also made 

 several other mistakes. The ventral cara- 

 pace came in contact with the lower part 

 of the "clavicular'" and was nearly parallel 

 to its narrow lower part. In consequence, 

 the distance between the ventral and 



dorsal part of the body was too short and the whole fish too flat. The size and position 

 of the antero-dorso-lateral were also incorrect, and the postero-lateral was not figured at 

 all. Nevertheless, Branson's reconstruction is one of the best, for it is based on fragments 

 of only one specimen. Therefore the relation between the si?e and position of the single 

 plates is more or less correct. 



The next year, 1909, Dean gave a short description with six photographs of a 

 reconstruction of Dmichthys terrelh, Nwb. (Text-figure 7), which is now on exhibit in 

 the American Museum of Natural History. This is the most complete reconstruction of 

 Dinichthys known to me. ' Unfortunately, in making it, plates from four different 

 specimens were used and some of these were made of plaster. Therefore the re- 

 lation between the single plates is not entirely correct. As a whole this reconstruction, 

 was very which good for its time, shows many mistakes: 



1. The entire carapace is too flat 

 and too broad, especially the head shield 

 and the median-dorsal plate which are not 

 sufficiently curved. 



2. The ventral shield is not related 

 to the dorsal shield. This circumstance 

 alone makes the reconstruction improbable. 



3. The clavicular (made of plaster) 



is too big in relation to other plates. The 



spinal and intero-lateral are not shown. 



T,, , ^. '^"^ T-^.' . , , „. They are included in the clavicular, and 



Deans reconstruction of Dmichthys terrelh 



Nwb. (1909). ^^^ incorrectly placed. 



ADL, antero-dorso-lateral; CI, clavicle; MD, median-dorsal; ' '^ 



PDL, postero-dorso-iaterai; SO, sub-orbital. Later the reconstruction was changed a 



lAccording to Dean, two other Museums (Harvard and South Kensington) have made attempts to reconstruct Dmichthys, but 

 the present writer has been unable to find anything about these two reconstructions in the literature. 



