280 C. M. CHILD 



/, g; of various degrees and kinds of union or incomplete separation 

 in h-s. These unions or connections of the eye spots are the most 

 frequent of the eye abnormalities and find their extreme type in 

 the single eye spot situated in the median line (fig. 5, fig. 6, r). 

 Differences of position and size of the pigment spots (fig. 6, 

 a, b, c,) are not uncommon but, as noted by earlier observers, 

 are sometimes, though by no means always, merely a consequence 

 of an oblique cut or of other incidental conditions which delay 

 the development of one side of the head, as compared with the 

 other. The development of more than two pigment spots or 

 complete eyes is of rather rare occurrence under the usual envir- 

 onmental conditions, but as will appear in a later paper, it can 

 be induced experimentally. In general, the characteristic abnor- 

 mality of the eyes which indicates the first visible approach 

 toward incompleteness iji head formation is the formation of 

 eye spots nearer together than in normal animals: between this 

 condition and forms with a single median eye all gradations 

 exist. The various degrees in this change from the paired sym- 

 metrically placed eye spots are in general parallel to the decreas- 

 ing capacity of the piece for head formation. In this and later 

 papers various factors which influence the development of the 

 eye spots will be considered. 



' Headless tails. ' The failure of certain pieces of Planaria to 

 reproduce heads has been noted by various observers, but so 

 far as I am aware, no one has stated correctly the facts concern- 

 ing their occurrence. Morgan's interpretation of the absence 

 of a head in these pieces as the consequence of the method of 

 closure (Morgan, '98, p. 380) is incorrect, as I shall show. The 

 head does not fail to develop because the wound closes in a certain 

 way, but the wound closes in a certain way because the head fails 

 to develop. The amount of new tissue formed at the anterior 

 end varies greatly in these headless pieces: the anterior wound 

 may remain widely open and the earlier stages of regeneration 

 may be identical in appearance with those in pieces where a head 

 develops. Such a case is shown in fig. 7. On the other hand, the 

 anterior end of a headless piece may appear as in fig. 8, i. e., the 

 edges of the wound may be closely approximated and but little 

 new tissue formed. But the analytic study of these pieces shows 



