100 MR. MIERS ON SEVERAL INSTANCES OF ANOMALOUS DEVELOPMENT 



by its pressm-e a corresponding groove close to the periphery of the albumen : the 

 maroins of the large cavity in the albumen, in which the much smaller embryo is 

 suspended, also correspond with this line of circuit of the raphe, so that the albumen 

 alono- this line is reduced to a very thin substance, and consequently is nearly separated 

 into two parallel slabs. In Stemonurus I could not perceive the indication of any 

 chalaza at either extremity of the inner integument, and there exists a more intimate 

 union of this uitegument with the testa, which are both of delicate texture. In Fennantia, 

 on the contrary, a distinct chalaza is ^dsible at the cotyledonary extremity of the inner 

 integument at the vanisliing point of the dorsal raphe : the same exists in Mappia, and 

 probably in aU other genera of the Icacinacece. Mr. Bennett (Plant. Jav. rar. 131) 

 quotes the observations of Mr. Brown and of M. Brongniart to show that the eversion of 

 the raphe upon the dorsal face of the seed is of frequent occm-rence in Celastracece, 

 Rhamnacew, AqnifoUacece, and CaprifoUacece, and is fully confirmed by his own experience 

 in the seeds of Euonymus and Bhamnus, where it is frequently everted, though often only 

 laterally displaced ; and he entirely adopts the opinion of Mr. Brown, that this eversion of 

 the raphe in no way militates against the universal law of anatropal development ; for on 

 examining these occui'rences at an eaidy stage of their growth, he invarialjly found the 

 ovule with its raphe on the ventral side, or that next the placenta, and he observed that 

 its subsequent appearance on the opposite or dorsal face, is due to an evident torsion of 

 the upper extremity of the raphe, or short fimiculus, during the increment of the seed, 

 and not to any original development. The eversion of the raphe in the Icacinacece, 

 together with a similarity in other essential points of structure, strongly prove the close 

 afiinity of this order to the Celastracece, AqnifoUacece, and other families which I proposed 

 to unite with it into one general alliance (the Dri/ales), nearly allied to the Rhamnacece. 



In the above-mentioned structure in Stemonurus, a cui'ious deviation from the usual 

 development of anatropous seeds occurs. In the organization of most of its parts there 

 exists, as before mentioned, a perfect analogy with that found in Fennantia : the seed is 

 albuminous, and the radicle of the embryo is superior and pointed to the hilum ; but while 

 in Fennantia we trace the raphe in a straight Une to the opposite or cotyledonary extremity, 

 as in ordinary anatropous seeds, we find in Stemonurus, owing to the singular retroversion 

 of the raphe (that is to say, its continuation beyond the point where it ought to have 

 tei'minated, and its subsequent ascent to the summit — to the point of its disappearance 

 near the hUum), that this imusual cu-cumstance is not attended by a corresponding 

 change in the position of the embryo, and offers a paradox difficult of explanation. In 

 ordinary cases, the ovule makes a semi-gyration on its centre, thvis producing a complete 

 inversion of the embryo and its integuments, all moving together in accordance : but do 

 we not see, from the course of the raphe in Stemonurus, an indication that a complete 

 gyi'ation or double anatropal action of one or both of the integuments has occurred, while 

 the embryo at the same time has effected only a half-gyration, or simple anatropal move- 

 ment ? How can we otherwise account for the peculiar form of the raphe in this case ? 

 I will recur again to this point when I come to speak of other developments which appear 

 to offer much analogy with this phoenomenon. 



In order to prepare the way for this investigation, it is of some interest to determine 



