° 1915 J Johnson, A Four-ivinged Wild Duck. 477 



number from the pentadactyl type of its class, the full number of 

 five digits should abnormally occur, these accessory digits might 

 in reality represent a reversion to the ancestral type ; as for example, 

 when a fifth finger occurs in some urodelous amphibians which 

 normally possess four fingers. 



A second view is that of double embryonic anlagen. Here the 

 normal anlage has become divided either through some extrinsic 

 perhaps mechanical agency, or through an intrinsic peculiarity of 

 the germ -plasm. 



According to a third view, the supernumerary digits or limbs are 

 simply malformations or pathological growths that belong in the 

 category of duplicate formations (Doppelbildungen) which first 

 arise as germinal variations, and are inheritable. 



In the efforts of the various authors holding the views just men- 

 tioned, Barfurth finds a more or less evident tendency to assign all 

 cases of supernumerary digits etc. to a common cause. He, himself, 

 believes that they result from a variety of causes. 



Among external influences the amnion is considered by some 

 authors as the cause of accessory appendages. Tornier ('97) con- 

 siders it an established fact that amniotic folds or bands are respon- 

 sible for some cases of supernumerary digits or limbs in mammals; 

 that this is true not only where such parts occur on one side of the 

 body, but also where they appear on both sides, similar and simul- 

 taneous. He cites the case of a pig's foot in the Zoological Institute 

 of the University of Leipzig, in which he declares one may follow 

 out in detail the history of the processes by which the end result 

 was produced. According to his view an amniotic band or fold 

 may press against the pelvis or a shoulder blade of the embryo in 

 such a way that a portion becomes pinched off; or a swelling or 

 protuberance arises in which a process of regeneration sets in, pro- 

 ducing a structure that in greater or less degree is a duplicate of the 

 part from which it sprang; or a growing limb bud may be split by 

 the penetration into it of such folds or bands. Tornier based his 

 conclusions upon a study of both birds and mammals. 



Opposed' to this view in regard to the influence of the amnion 

 stand the observations of Kaufmann-Wolf ('08). In an extensive 

 study on the domestic fowl in adult and in embryo, this investi- 

 gator found no evidence that the embryonic membranes, amnion 



