THYNG: SQUAMOSAL BONE. 
403 
confirmation from the observations of Case (: 05). In some (Amphi- 
chelydia, Cheloniidae, Dermochelydidae) there is a complete bony roof 
in the temporal region resembling considerably that found in the 
Cotylosaurs, except that instead of three, only two elements occur. 
Baur (’89), reasoning from this condition, attempted to explain the 
structures existing in other Testudinata by supposing that supra- and 
infratemporal fossae were never developed in the order, but that the 
original bony roof suffered diminution by absorption from behind, 
between squamosal and parietal (e. g., Chelydridae, Dermatemydidae, 
Staurotypidae, Cinosternidae, Platysternidae, Emydidae, Testudin- 
idae, Trionychia). But, simultaneously with this absorption, a reduc¬ 
tion or loss occurred from below between quadrate and jugal, result- 
ingin the formation of a slender arch which in a few cases (Terrapene, 
Geoemyda) was entirely reduced. In other cases reduction took place 
entirely from below leaving the squamosal and parietal either in 
contact as in the Chelydidae, or destroying the connection as in 
Chelodina. 
Plausible as this may be, it is by no means proven, and to-day there 
is great uncertainty as to the homologies of the two bones of the Testu- 
dinate temporal region. The more ventral element called by Hall- 
mann the quadratojugal (temporal ecailleux of Cuvier; squamosal, 
Owen, ’66- 68) and so regarded by the majority of authors, has been 
called the paraquadrate by Gaupp (’94) and thus homologized with 
the bone in the Stegocephala and Cotylosauria which lies dorsal and 
medial to the true quadratojugal, a view which has been followed by 
Siebenrock (teste Rabl), Wiedersheim (: 06), etc. The more medial 
element (mastoi'dien, mastoid, of Cuvier and Owen) is usually termed 
the squamosal and thus homologized with that bone which in the 
Cotylosaurs I have shown to be the supratemporal. Baur (’89, p. 
472) using the nomenclature of the present paper, held that the 
turtles lack a supratemporal bone while Williston (: 04) thinks that 
the squamosal (his prosquamosal) is absent, its function being taken 
by the quadratojugal. 
The questions involved can only be settled by study of the develop¬ 
ment of the bones of this region or by the discovery of more primitive 
types of Testudinata. 
Rhynchocephalia .— Before considering the skull of Sphenodon it is 
best to consider some of the more primitive members of this order, 
where, although exact knowledge of the temporal region is not abun- 
