1891.] MICROSCOPICAL JOURNAL. 8 



to represent the diatom cell by the use of paper models or other ol)jects. 

 One of the best I know of is a cigar-case made of two halves, one of 

 which slides over the other. It is a much better object for the purpose 

 than the pill-box for such forms as Isthmia and Jyiihiulphia. It is 

 capital for these, and also, when viewed endwise, for the Navictda 

 forms. But two paper cylinders with convex ends, one cylinder en- 

 closing the other, are perhaps the best of all, as being the most pliable 

 and as representing the typical form of the diatom, which is that of a 

 cell. When the two valves are prolonged the diatom cell becomes a 

 cvlinder ; when the cell is compressed it takes the oval or boat form, 

 or the form of a triangle or square, as the case may be. Now. paint 

 a stripe around the open end of each half of the paper cylinder, 

 where one slips over the other, and you have the true representation 

 of the hoop, and will never be puzzled again about its relation to the 

 valve, and when you draw IstJunla in front view you will not draw the 

 edges continuous but broken, showing that one valve is partly inside 

 the other. 



i\fter you have thus made yourself thoroughly acquainted with the 

 true shape of the diatom cell and the relation of its two valves, get 

 hold of Wolle's Desmids and study the plates carefully. Compare the 

 front and the end views of the same desmid and see how remarkably 

 they difler, and you will understand still better the difiei-ence between 

 the front and side views of the diatom, for the end view of the desmid 

 corresponds to the side view of the diatom. And you will also note 

 the close correspondence between the diatoms and desmids as a whole, 

 and have a fixsh proof of the vegetable nature of the diatoms and of 

 the correctness of the place which has been assigned them in the vege- 

 table world among the algic and close to the desmids. Look at the 

 filamentous forms of the desmids and see how much they resemble the 

 filamentous forms of the diatoms. And then study the growth of the 

 desmids by division, and you will be able to grasp the true idea of the 

 growth of the diatoms by the same process, for there is the same 

 dividing of the one cell into two by the formation of two new semi-cells 

 between the halvesof the mother-cell. See pi. xiv, figs. i3, 13, WoUe, 

 of a Cosmarium so dividing and multiplying. See also pi. xxii, fig. 

 ^, 'iox XatitJiidijitn in process of division. Also pi. xlv, figs. 19, 20, of 

 Staiirastriim. And note particularly that the new 'semi-cell is at first 

 smaller than the semi-cell from whicli it grows. And the reason why 

 it afterwards reaches the full size is because there is no prolongation of 

 a silicified membrane enclosing and limiting it as in the case of the 

 diatom. And that is one of the distinguishing features between the 

 diatoms and the desmids. In the desmid the two halves or valves are 

 joined edge to edge, are exactly alike. In tlie diatom on the contrary, 

 or at least in most cases, the two halves or valves are joined like two 

 joints of a telescope, so that one half is always just large enough to 

 slide over the other half. And as the free, open end of each half of 

 the diatom is silicified as well as the rest of tiie valve, the semi-cells 

 formed within are necessarily smaller than their respective mother-cells 

 or original semi-cells. But while one of the new frustules will be 

 smaller than the frustule from which it is produced, the other will be, 

 or may be, just as large. Your paper model will show this. For it 

 is evident that another valve can be formed inside the lanjer of the two 



