228 THE AMERICAN MONTHLY [October, 



Imbeddiii!? Seeds by the Paraffin Method.^ 



By W. W. ROWLEE, 



ITHACA, N. Y. 



The modifications that may be made of the paraffin method of imbed- 

 ding objects for sectioning, are very many. There is always, however, 

 some danger of shrinking delicate and very soft plant tissue. .This is 

 due to the use of heat in the process of infiltration ; and probably some 

 of the non-heat-employing methods will be found preferable where such 

 delicate tissue is to be imbedded. But, for objects that will withstand 

 this process of infiltration, the paraffin method has many advantages 

 over others. Imbedded in paraffin, objects are held firmly and may be 

 preserved as long as desired without further attention. 



For imbedding mature seeds I have found nothing equal to paraffin. 

 The texture of the seed is often very dense, and offers much resistance 

 to the knife. For this reason I found it better to use the harder grade 

 of paraffin. A second serious difficulty that was met with in imbed- 

 ding seeds, was the fact that there was little, if any, tissue connecting 

 the embryof with the seed-coats. Thus it would happen too often that 

 just as the sections were being taken through the middle of the seed, 

 and the most valuable ones are those near the centre, the embryo would 

 leave the coats and the whole series would be spoiled. The inner sur- 

 face of the inner coat in many seeds is highly polished, and as soon as 

 there is nothing to retain the embryo but its adhesion to the coat, it will 

 loosen. The paraffin does not hold the two together as would be ex- 

 pected. It was suggested that, in order to soften the tissue and thereby 

 make it more susceptible of infiltration, it would be well to thoroughly 

 soak the seeds in water before hardening in alcohol. This was tried, 

 and there was a great improvement in the results. Fewer of the sec- 

 tions went to pieces after they were transferred to the slide, and the 

 parts of the seed kept their respective positions much better. 



In order to study the microscopic structure of seeds, much more sat- 

 isfactory results can be obtained if the sections are kept in series. It is 

 often necessary to have two or more successive sections before a correct 

 idea of the seed can be obtained. 



The method is a modification of the one used and taught in the his- 

 tological laboratories of Cornell University. In its practical application 

 it is as follows : In choosing seeds to section great care is taken to get 

 those which are well filled. This precaution is especially important, 

 as many seeds, for various reasons, never develop more than the coats 

 or the enveloping ovary coats. If a seed has a straight embryo or even 

 a bent or curved one it is better to determine by dissection just how the 

 parts of the embryo are arranged with reference to the external parts 

 of the seed. Thus the seeds of Helianthus tuber osUs are flattened 

 and slightly wedge-shaped. The embryo w^ithin is straight and the 

 upper or inner surface of the cotyledons lie in a plane parallel to the 

 plane in which the seed is flattened. Moreover the cotyledons are in 

 the upper broader end of the seed. Where the seed has no external 

 character, as in a eupatcorium, by which the position of its internal 

 I . 



* Read before the American Microscopical Society at Detroit, August, 1890. 



t The term " embryo " is used here where 011 som ; accounts it would lie better to use the word " nu- 

 cleus." The embryo is often but a very small part of the substance contained within the seed-coats. 



