224 THE AMERICAN NATURALIST. [VOL. XXXIX. 



(:O2, pp. 167, 168) seems justified in emphasizing their resem- 

 blance to pseudopodia rather than to any other structure of the 

 cell. If they should finally be connected by intergradations 

 with the exceedingly fine plasmodesmen of Strasburger, there 

 would stand at one end of the series structures so thick as to 

 be composed of a plasma membrane containing much cytoplasm 

 in the interior and behaving like haustoria or pseudopodia and 

 at the other end delicate fibrillae. Viewing the problem of their 

 relationships from the lower plants upwards, it is very difficult, 

 if not impossible to follow Strasburger's theory that all cytoplas- 

 mic connections (plasmodesmen) are related to developments 

 from the plasma membrane similar to cilia. They seem to be 

 more of the nature of processes put out from the cytoplasm and 

 when necessary penetrating cellulose walls probably in response 

 to chemotactic stimuli since they are most conspicuous when 

 metabolic activities are obviously important (e. g., nourish- 

 ment of the egg in gymnosperms and sporophytic generation 

 of the red algae). 



In method of development we have seen that protoplasmic 

 connections fall into two classes: (i) those that represent the 

 incomplete separation of daughter cells, and (2) those that 

 result from the coming together or fusion of protoplasmic out- 

 growths. The types of the first group are always in the be- 

 ginning open communications which later may become largely 

 or wholly closed ; types of the second group may result in broad 

 cytoplasmic fusions (e. g., many fungi) but there is evidence 

 that in many cases, especially among the higher plants, the two 

 processes only come in contact so that the plasma membranes 

 are applied to one another but do not actually unite. It does 

 not seem probable that the two methods of development or the 

 presence or absence of intimate protoplasmic union indicate a 

 different kind of structure. They are more likely to be only 

 varied responses to the demands for a more or less close associa- 

 tion of neighboring cells. Broad communications are especially 

 characteristic of regions where there is evidently an extensive 

 demand for the nourishment of a cell or tissue, as in the eggs of 

 the cycads or the cystocarp of the red algae. 



The functions of protoplasmic connections are probably vari- 



