1901) MICROSCOPICAL JOURNAL. 129 
but if they were coarse they were lost sight.of by virtue 
of their largeness. 
ComMENTs BY E. M. Netson.—Mr. Merlin used the tel- 
escope, till his eyesight was exceptionally keen, which was 
probably as good if not better training for the eye than 
the microscope. The fact of one being unable to see any 
particular structure described in this paper would not, 
therefore, be evidence that Mr. Merlin was likely to have 
been mistaken in what he had seen. He had tried a 5-6th 
cone with the dry 4 mm. apochromat of ‘95 N.A. with the 
34-inch wick of a paraffin lamp and an acetate of copper 
filter, but was not able to effect resolutions to anything 
like the extent Mr. Merlin had done. He next tried sun- 
light with a-heliostat, but the heliostat proved untrust- 
worthy and the sunlight fickle, so he was not able to push 
his experiments as far as he would have liked. He found, 
however, that with sunlight he could use a filter of much 
greater thickness, and then he was able to see some of the 
structures. There was another point—viz., that the Abbe 
diffraction theory did not fit in with all the observed phe- 
nomena bearing upon that branch of microscopy. It was 
highly probable that the large solid axial cone had a 
greater resolving power in it than was generally sup- 
posed. His experience showed him that 80,000 times the 
N.A. of the objective was the resolving limit in inches 
with this kind of illumination, but from what Mr. Mer- 
lin had said it was evident that a larger coefficient must 
be employed. The little beads in the lines on the hoop of 
a Pinnularia major were, so faras he knew, unresolva- 
ble by oblique light, but with the 5-6ths solid axial cone 
he had been able to see them with the dry 4 mm. apo- 
chromat. Strange to say, this same object had in 1895 
been a kind of minimum visible or crucial test for an apo, 
4” of 1:43.N.A. It appeared, therefore, that the “mini- 
mum visible,” the “crucial test,” the “scarcely resolvable 
detail” of one year became the commonplace object at a, 
