THE BINOCULAR MICROSCOPE. 433 



^because it deals with the coarse features of a fairly large 

 diatom and not with periodic structures, such as striae, secondary 

 markings, etc. As a test with striae a moderately fine Navicula 

 ■lyra may be used. The Parallel tube instrument will resolve the 

 transverse striae when the valve is in any direction, while the 

 Wenham will only do so when the valve is vertical, or nearly so. 

 ■ The stereoscopic relief in these experiments was stronger in the 

 Parallel tube binocular. 



Let us now examine a Coscinodiscus radiatus, which is an excep- 

 tionally flat diatom. The Parallel tube binocular revealed a 

 feature never before observed, viz. that the "Omphalos" is at 

 the bottom of a small pit. Some other diatoms were then 

 examined with reference to this detail, and some other Coscinodisci 

 were found to have similar pits, an Arachnoidiscus ornatus from 

 Japan had one also. The diatom was then examined under a 

 similar power (270) in the Wenham. Knowing that this little 

 pit existed, its presence could be recognised, but it is doubtful 

 if it would have been discovered if its presence had not been 

 known. The fact that it has never been previously seen proves 

 that this is so. 



The Parallel tube binocular has revealed another feature of this 

 diatom, viz. that its floor is anything but flat ; parts are raised and 

 parts depressed, this quite irregularly and not in any symmetrical 

 pattern. By way of illustrcition, suppose some pastry dough had 

 been spread smoothly on a table, and that some children had 

 ■ come when the cook was not looking and had pressed in portions 

 of the dough with their fingers and hands. If a layer of empty 

 honeycomb cells could have been treated in the same way as the 

 soft dough, we should have a fair representation of the surface 

 of this diatom, which hitherto has been thought to be flat. The 

 spread slide, which contains this diatom, was bought in the year 

 1868; it has been kept fifty years in a special box of test objects, 

 and not in the diatom cabinet. It is quite impossible to say how 

 many hundreds of objectives have been tested on it, but probably 

 there is nowhere a diatom slide that has been worked over to 

 anything like the same extent as this one. Imagine then my 

 -surprise on seeing quite a new aspect of this well-studied 

 <^C. radiatus. It was as great as if I had found the hours on my 



