Further Xotes on Seed Structur* 8. 137 



" In the apical region of the seed the testal layers were found 

 most difficult to delimit owing to a lack of sections traversing the 

 long and slender micropylar tube in the true median longitudinal 

 position, and the large amount of pvrite investing the peripheral 

 tissues of the strobilus in irregular areas." 



In no form are the testal features conserved with more of 

 diagrammatic clearness than in the French specimen Bennet- 

 tites Marierei, Lignier' s description of which soon followed 

 that of Solms as already mentioned. In fact the seeds of this 

 isolated strobilus are conserved in such unrivalled perfection 

 as to leave nothing more to be desired except the recovery of 

 a second example, and make us anxiously ask if French col- 

 lectors have really sought for further material in the original 

 locality as diligently as they might. From the description of 

 Lignier (1S9A) which enters into much detail we need not, 

 however, quote in full in the absence of a series of photomicro- 

 graphs. Though before passing on to the American seeds it 

 should be mentioned that the limiting region between the 

 seeds and surrounding scale-formed seed pit, which as we have 

 just seen Solms found difficult to see in B. Gibsonianu.s, is 

 always distinct in B. Mbrierei. In any of the sections from 

 the latter one may clearly see that the appearance, as of a 

 peripheral confluent growth of the seed wall with the inter- 

 seminal scales, is due to appression and the failure of the scales 

 in places to develop their epidermal layer continuously. 

 There thus arises & pinching out into a single ribbon-like layer 

 of interseniinal epidermis appressed closely to the very outer- 

 most cells of the seed. The condition does, however, show 

 how. as the result of close growth and suppression of epider- 

 mal cells, a form of intergrowth could easily arise. Quite the 

 only structural feature of the testal wall that Lignier leaves in 

 doubt is the nature of the middle layer, whether stouy, or. as 

 he is inclined to believe, fleshy. On this point it is. however, 

 reserved for the American species to throw a clear light, and 

 to comparisons with them we may now turn. 



Comparison of testal structure in the American and European 

 Cyiadeoidean seeds* 



On comparing the descriptions of the testa already quoted, 

 namely those of Carruthers (1), Solms (2), and Lignier (3), 

 with each other, and with those given for the American seeds 

 by the writer, a most substantial agreement within strict 

 generic limits will be found and the differences due to conser- 



* In these comparisons the names of testal features and seed parts proposed 

 by Oiiver and Salisbury (11) are employed, having been found convenient 

 and nsable. while a close reading of the present text will disclose one or two 

 extensions of this necessary nomenclatu re of use. 



