248 PROCEEDINGS OF THE ACADEMY OF [^Mavcll, 



as the reraaiuing parts are concerned, they present the same frag- 

 mentary character as in the preceding stage, being in certain loca- 

 tions almost unrecognizable. I noticed in the present series (and 

 likewise in several later ones) that there is no necessary corre- 

 spondence either in the number, length or distribution of the frag- 

 ments of the two sides. In the case of the specimen of the present 

 stage examined the sections were almost exactly transverse, so that 

 the same parts were cut on both sides. Yet the tubal cord may 

 be present for a considerable distance on one side and apparently 

 altogether absent on the other. This irregularity is a marked 

 feature during the entire metamorphic period. I find that there 

 is also marked individual variation in this respect. This vari- 

 ability would seem to indicate that the character of the fragmen- 

 tation is not due to some inherited tendency, but is produced by 

 mechanical forces exerted by the surrounding structures. 



Posteriorly the Eustachian cokI terminates suddenly in the usual 

 position, dorsal to the anterior extremity of the " lateral recess " 

 of the branchial portion of the pharynx. In the same region a 

 prominent proliferation arises from the dorsal wall of the " recess," 

 and extends upward to the same relative position as that occupied 

 by the cord in the more anterior sections. This structure may 

 represent the same mass of cells which originally established the 

 connection between the tubal cord and the wall of the pharynx, but 

 of this interpretation I am uncertain, since I was unable to discover 

 any sign of such proliferation in the preceding stage or in a num- 

 ber of later stages. Possibly its occurrence or absence is a matter 

 of individual variation. 



There has been but little change in the skeleton since the last 

 period. Posteriorly, however, the quadrate has developed a pos- 

 teriorly projecting processus oticus, which comes in contact with 

 the ventral surface of the auditory capsule. The processus oticus 

 arises at the angle formed by the body of the quadrate with the 

 processus ascendens. The stapes appears for the first time as an 

 oval chondrification within the membrane closing the fenestra 

 ovalis. There are no distinct traces of a columella auris. 



Stage VIII. — Tadpole of 21 mm. Preceding the appearance of 

 the fore-limbs. 



At this time we have the earliest distinct^ppearance of the annu- 

 lar cartilage (PI. IX, lig. 28, An.). About opposite the point where 



