95 



consist of two convex shields, with two or perhaps four descending 

 points; one is placed above the other, and the convexity of both is up- 

 ward. The knobbed foot is evidently jointed to the under surface of the 

 lower shield, for the whole upper part is capable of considerable mo- 

 tion from side to side, as on a pivot, while the foot is steady. The 

 foot reaches to the very bottom of the gizzard, and the end which in 

 one aspect appears dilated, in another seems to have a curved spoon- 

 like extremity. 



The mallei, viewed in situ, appear each to consist of a rod, hang- 

 ing freely down in the gizzard, nearly parallel to the incus, but 

 slightly bowed outward, and of an arm or head, set on its upper end, 

 nearly at right angles. The end of this arm appears cut into teeth, 

 which meet those of the opposite one upon the surface of the upper 

 shield of the incus. A figure of the apparatus thus viewed is care- 

 fully drawn at G ; the dotted lines representing the position into 

 which the mallei are alternately thrown in the process of mastication. 

 The specimen from which this figure was taken afforded me an 

 unusually good observation. On putting carmine into the water, and 

 examining the animal carefully with a power of 300 diameters, the 

 jaws worked slightly, the points opening a little way, and then clos- 

 ing; the rods of the mallei were drawn towards the bottom for 

 opening, and upwards (as in the dotted lines) for closing. A little 

 mass of pigment was soon accumulated beneath the tips of the jaws, 

 which spread itself over a rounded surface (fig. G), but did not 

 pass farther, nor did an atom at this time go into the stomach. 



On applying pressure by means of Mr. Boss's compressorium, the 

 structure of this apparatus becomes more clearly resolved. From 

 many examinations, I have found that the rod of the Tnalleus (fig. E) 

 is a long loop of bone (or a substance analagous to it), one side 

 of which only (the interior) is attached to the head-piece, which, 

 nevertheless, has a descending process that nearly meets the outer 

 side of the loop, and is, I think, connected with it by the soft part. 

 The head-piece has also another process, descending obliquely be- 

 hind the loop. I have distinctly seen floating bands (muscular ?) at- 

 tached to the bottom of the gizzard by one end, and by the other to 

 various parts of the bony frame -work. One pair of these proceed up to 

 the oblique hind processes (one to each), to the extremities of which 

 they seem attached : another pair go from the foot of the table to the 

 curved points that project on each side ; their attachments to these 

 points I saw very clearly : another pair less clearly appeared to pass 

 up to the higher parts of the table between it and the loops. All 



