52 



JOHAN KIÆR. 



M.-N. Kl. 



In conclusion, we will briefly examine the family relationship between 

 the two families here dealt with, and other Lower and Middle Cambrian 

 forms. We must first of all consider the genus Agnmlos Corda. This has 

 always been regarded as a very near relative. Indeed Strcmtella forms 

 were formerly described as belonging to this genus ; subsequently they 

 were understood to form a sub-genus, or a genus that was near to Agroulos. 

 With Ellipsocephalus the}- would constitute the family Ellipsocephalidae. It 



seems to me, however, that the accurac}- of this 

 view is somewhat doubtful. 



There is undoubtedly considerable likeness 

 between these genera as regards a number of 

 structural features, e. g. the ordinary form of 

 the bod}' and the structure of the segments. 

 As pointed out by Barrande, the latter in 

 particular, with its markedly geniculated and 

 remarkably furrowed pleurae, appears to indicate 

 closer relationship. At the same time it has not 

 been sufficiently emphasized hitherto that the 

 cranidia exhibit a marked dift'erence in structure. 

 Whilst Streiinella and Ellipsocephalus are 

 specially noteworthy for the fact that the palpe- 

 bral lobes are large and extend far backwards 

 to the occipital furrow, and that the facial suture 

 behind these has a quite short and reflected course, Agranlos on the 

 contrary has small palpebral lobes that are situated farther forward, and 

 long branches of the facial suture extending outwards (fig. 7). These give the 

 cranidium quite a difterent character, and possess additional significance 

 because we cannot demonstrate an earlier transition from the Stremtella type 

 to the Agranlos type. The latter was already fully developed in the Lower 

 Cambrian age. From this we have Agranlos Redpathi Walcott\ and 

 Agranlos Stator Walcott-, recently described by the same author. The 

 later, with its 22 segments, is a very primitive form,, the systematic position 

 of which may be somewhat doubtful. The genus thus continues in profuse 

 development in the Middle Cambrian, where several remarkable groups and 

 series of forms may b}' distinguished. Thus we have the ceticephalus 

 group with the type of the genus Agranlos ceticeplialns Bark, and several 



Textfig. 7. Cva.n\à\a.oi Agranlos 

 ceticephalus (a) and Ellipsoceplui- 

 lus Hoffi (b) after Barrande. 

 Shows the difterence between 

 the two tj'pes. 



1 Ch. D. Walcott. Fauna of the Olenellus zone, p. 654, fig. 69. 



2 Ch. D, Walcott. Cambrian Geology & Palaeontology, 111, No. 3, Cambrian Trilobites 

 iSmithson. Misc. Coll. Vol. 64, No. 3, 1916, p. 173, pi. 36, fig. 6). 



