1- JOHAN KiMR. [sec. arct. exp. fram 



In the examples of Mncrop. Scheli before us we see on the right 

 side that the post-orbital canal at the liinder part bends shghtly inwards 

 and tapers. I cannot see, however, that they then proceed downwards 

 with funnel shaped openings as in the case of rapheidolabis. On the 

 other hand we can see that from the hinder point of the occipital canal 

 a new canal runs with a shghtly arched direction outwards and 

 rearwards to the hinder margin of the cranial roof. In my opinion 

 this appears to be the beginning of the lateral line canal. As 

 is well known a similar condition is seen in the case of a number of 

 coccosteomorphous fishes, and is therefore a further proof that the entire 

 head shield of Macropetalichthys contains only the cranial roof. 



The sensory canals which have a breadth of about 1 mm. are 

 very clearly discernable in the specimen before us, owing to the circum- 

 stance that the bone substance has almost entirely fallen away. As is 

 well known, the canals in this genus are so deeply embedded in the 

 dermal plates, that in the form before us they appear on the under side 

 of the cranial shield like ridges. They open upwards by means of a 

 series of pores, which appear to be found only on the outer side of 

 the canals. They are easily discernable along the outer side of the 

 right post-orbital canal and for a short distance along the right supra- 

 orbital canal, mside the orbits. At the former place they are round, 

 with a cross section of 0.3 — 0.4 mm., and at the latter, where the entire 

 bone substance has been preserved, thev are oval, about 0.4 — 0.-5 mm. 

 in length, and 0.1 mm. in breadth. 



The specimen also clearly shows the impression of an extensively 

 branched and fine canal system on the inner side of the cranial roof 

 (pi. I, fig. 4). As far as I can understand these fine canals do not 

 appear to have had anything to do with the sensory canals, but may 

 possibly be veins that have been embedded in the cranial bones. I have 

 not been able to find anything corresponding to them in description of 

 the other forms known of this genus. 



Orbits. These are in the main discernable, although the foremost 

 portion of the cranium is defective. On the left side the length mav 

 be determined at 11.5 mm.; it is more difficult to determine the breadth, 

 but presumably it may be placed at 7.5 mm. The orbits may therefore 

 be reconstructed as shown in the drawing, fig. 2. They thus appear 

 to have been unusually large and lengthy. 



Horizon. From the zone Dh in marin facies. Goose Fiord, 

 Ellesmere Land. 



