14 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 85 



slightly constricted at the union of the trunk segments and extends 

 back to the anal segment which protrudes posteriorly from beneath 

 the shield ; the gullet connecting the mouth and the stomach must have 

 extended forward and upward. 



The large hepatic glands and caeca are somewhat similar to those 

 of Burgessia bclla (text fig. 5) ; a short, strong tube with a well- 

 marked anterior and posterior tube leads out from the stomach, and 

 branching from these lateral tubes are a series of hepatic caeca. The 

 small hepatic caeca are located between the long posterior tubes of the 

 hepatic glands and the intestine and posterior to the main hepatic tube. 



Owing to the excellent state of preservation of many of the speci- 

 mens showing the hepatic caeca it is probable that they were situated 

 between the dorsal and ventral membrane of the carapace and thus 

 held in position and protected from destruction ; that they are pre- 

 served at all is one of the wonders of this remarkable Burgess shale 

 fauna. 



Functions of appendages. — These were presumably the same as 

 for similar organs in Marrella and Burgessia, and the mode of occur- 

 rence is essentially the same. 



Diagraminatic restorations. — The diagrammatic restoration (fig. 2) 

 presents the outline of the carapace and posterior shield with the 

 stomach, intestine, hepatic tubes and caeca of the digestive system 

 outlined, also the thoracic trunk, the telson and the thoracic limbs 

 as far as known and interpreted. The data for this diagram are very 

 good except the jointing of the endopodites and the exact form of the 

 coxopodites and proximal joints of the exopodites. In figure i, the 

 endopodites on each side have been omitted and the exopodites drawn 

 in so as to show their structure and position above the endopodites 

 and below the ventral membrane of the posterior dorsal shield. The 

 sixth limb has both the endopodite and exopodite attached ; this should 

 be compared with the thoracic limb of Marrella splendens (text fig. 9). 



We know so little of the cephalic limbs of Naraoia compacta that I 

 hesitate to give a diagrammatic sketch of them, and it would not be of 

 even tentative value if we did not have the cephalic limbs of Marrella 

 and Burgessia for suggestion ; from the latter and from the evidence 

 afforded by a few specimens, the outline of figure i is drawn. 



Comparison with crustaceans. — Naraoia has many characters in 

 common with the trilobite and some in common with Marrella, Bur- 

 gessia, and Waptia, which will be spoken of in the discussion of this 

 group of genera. 



Plesiotypes. — U. S. N. M., Nos. 83945a-e. 



