No. 436.] APPENDAGES OE TREMA TASPIS. 



2 35 



Neither can I agree with Rohon's view that the oral surface 

 of the St. Petersburg specimen which shows the oral plates in 

 situ is so distorted as to force the median plate strongly to one 

 side. As I have pointed out elsewhere the true anterior median 

 oral plate is here almost exactly in its proper place, and it is very- 

 different in size and shape from the isolated one Rohon wrongly 

 supposed to be a median oral plate. 



The location of this shield shape plate in Tremataspis, for 

 there can be no reasonable doubt that it belongs to this genus, 

 has been a great puzzle. The objections to Rohon's view that it 

 is a median plate lying just behind the mouth are overwhelming. 

 It is (i) much too large. (2) It is asymmetrical and cannot be 

 regarded as a median plate. (3) Its beautifully preserved struc- 

 ture is totally unlike a jaw which we must assume it to be if 

 placed in that position. (4) The rounded collar at its anterior 

 end, the two powerful and counteracting muscles attached to its 

 inner side, and the sharply cut grooves on the polished outer 

 surface, that were unquestionably worn by contact with some 

 other hard part, shows conclusivelv that this plate was attached 

 by the collar on its anterior surface to some circular opening and 

 that it had a wide range of rotary movement from side to side, 

 an improbable movement in a plate thus located. 



The resemblance of this plate to the hypostoma of Trilobites 

 in its shape, muscle markings and its probable mode of motion, 



collar to the anterior ventral margin of the dorsal shield. Hut 

 its asymmetry, which at hist did not impress me so forcibly as it 

 did later, its large size, and the lack of correspondence between 



