194 Mr. Griffith on the Ovulum o/Santalum, 



was inclined to consider that a reduction of parts similar to that which con- 

 stitutes in my opinion the second modification of Viscum, occurred in Loran- 

 thus. But the obvious continuation of the tubes or sacs high up the stigmatic 

 canal seemed to present very obvious analogies with what is known of pollen 

 tubes. Besides this, the growth of the great bulk of cellular tissue, constituting 

 the first steps in the development of the embryo, could not be made to agree in 

 direction or situation with the similar growths in any modification of this sac 

 known to me. With the view of determining this point, I endeavoured in 

 many instances to trace the tubes upwards to the stigma, and, if possible, to 

 the pollen grains; but I did not succeed in tracing them more than half-way 

 up the style, nor is this particular species well calculated to promise success 

 from the length of the style. 



Although I do not see any absolute theoretical objection to the attachment 

 ot an ovulum to any part of the stigmatic canal, the inner surface of which 

 appears to me to have such direct relations with the placenta, , or to its being 



reduced to a simple membranous sac , yet the analogies were, I thought, in 

 favour of the denvation of (hese tubes from (he poi|en ^ ^ 



were' of! " ZT lT* ^ """* °" WMch ' f ° Unded this " °P™<» 



I io f an , T r ' WUh the eXCepti0D ° f that M "*•* the 



eU ved k. TTn '** ^^ ^ ° f wMch no ■—£ was, I 



S3ir; ZttzE in r icable a n,,e otherwise ^ 

 -^-^^^tssssr* that when an 



oe^rr::;:; £* rr-r appearins to exist before the 



to boyau* But theT ' " k ' ^ h 8trUCtUre and prance 



«c, which is d rived fo ln rh , '' Md bet ^» 'hatformof embryo- 



■nany things ^JZ ' 2^ £V£" «*»■ - *■» tuLs 

 are generally the innermost membranes of h ^ ^"^ ' ^ 



contents of both appear to iToT uT ^^ rtmctnre ' and the 

 «o Sa„ ta , um to point out h!r ts 7 *! """" a " d l ™ d «"7 *■*> 

 -e "and an ^,^2^^^^ ^ & *** 

 there are grounds for supposing that to e a ,h I * WWB embr y^cs, 



