1884-85-] Edinburgh Naturalists' Field Chtb. 245 



species, tlie variation in tlie number of stri;"o never exceeds one- 

 fifth ; and (4) that there is never any difference in number in 

 sj^ecies originating from the same stock. Dr Wallich ^ had pre- 

 viously, in 1877, expressed the general conviction that though 

 the total number of strife on the valves of a Diatom may be almost 

 constant in every valve of the same species, the number on any 

 fractional part of any valve (e./y., on the thousandth of an inch) 

 would vary in proportion to the size of the valve. 



To the opinion of Castracane just quoted both Kitton and Pro- 

 fessor H. L. Smith ^ have offered objections, and among these 

 perhaps not the least important, from a practical point of view, 

 is that pointed out by Smith — namely, the multiplication of 

 species which must follow its adoption ; although it ought always 

 to be borne in mind that, by basing their limits on broader 

 and firmer principles, their number should, if practicable, be 

 reduced, while the very diverse character of the stritB in Stauro- 

 neis gracilis, and its sporangia! frustule, Stauroneis phenicenteron, 

 is referred to as an extreme case proving the unimportance of tlie 

 deductions. Nevertheless in the description of any given frustule, 

 the position of the stria?, whether radiate or parallel, their moniliform 

 or confluent nature, and their occurrence over the general surface 

 of the valve or only over certain more or less definitely circum- 

 scribed areas of it, should be noted, as well as the presence or 

 absence of a median raphe, the existence or non-existence of 

 central and terminal nodules, and the figure and position of the 

 surface areote. 



So far the valves. The cingulum or connecting membrane can- 

 not be regarded as of essential importance, but it is often present 

 in some degree of pi'ominence. In circular and discoidal Diatoms 

 it presents the form of a continuous ring — e.g.^ in Coscinodiscus 

 radiatus, a form not nncommon in the littoral areas of the Firth 

 of Forth. In oblong frustules, again, such as Navicula, it has an 

 oblong or i^avicular outline. In not a few cases it presents an 

 elegant sculpturing either in the form of areolation or striation, 

 as in Isthmia and Achnanthes. In general, however, the propor- 

 tion of siliceous matter present in it is, as above noted, less than 

 in the valves. In size it varies very much, being very small in 

 Pleurosigma?, while in some instances it is distinctly larger at 

 one extremity than at the other — e.g.^ in Gomphonema. The 

 mode of its development, and of its behaviour during the temno- 

 genetio process, has not in all cases been clearly determined.^ 



1 'Month. Micr. Jour./ vol. xvii. p. 61 : 1877. 



2 ' Amer. Micr. Month. Jour.,' vol. ii. pp. 2-J1-223 : 1881. 



^ See, however, Flugel's results in 'Jour. Roy. Micr. Soc.,' 1884, p. 076 

 tt seq. 



