On the Structure of Diatoms. 23 



repeated, have convinced me that they are never both in focus at 

 the same time, and further that the longitudinal lines are on the 

 external surface, and the transverse on the internal plate. If there 

 are not two plates, the lines may be on opposite surfaces of the 

 same plate. 



White cloud illumination is found to be much better than other 

 and more brilliant light for demonstrating these slight differences 

 in focal distances. Many errors of interpretation are avoided by 

 using an approximately central pencil when the instrument used 

 is capable of elucidating all the details of structure without greater 

 obliquity. 



Frustulia Saxonica. — In addition to my observation of longitu- 

 dinal lines upon this test and resolution into dots,* it may be 

 worth noting that even with lamp illumination the ^V lias displayed 

 the transverse much clearer than they appear in Dr. Woodward's 

 photo-print.t With oblique light direct from a small German 

 student's lamp, without mirror, prism, or condenser of a-iy kind, 

 a person entirely unaccustomed to the microscope could distinctly 

 see them with Beck No. 3 eye-piece, power 7000 times. With 

 the ammoni-sulphate of copper cell the longitudinal lines and dots 

 are displayed with ease. 



This is one of the most difficult test diatoms thus far studied, 

 ranking but little easier than A. pellucida, N. crassinervis, and 

 Nitzschia curvula. 



Amphifleura pellucida. — Many times the writer has been able 

 to confirm the observations of longitudinal lines on this most 

 difficult test shell, but never has succeeded in seeing the dots except 

 with the blue cell and Wenham's paraboloid, and only then under 

 favourable circumstances.^ When resolution is effected, the dots 

 are exceedingly minute and uniform in size, showing as mere points 

 of light when magnified 2500 times. On one occasion the writer 

 has seen fine dark hues crossing between the transverse striae like 

 the steps of a ladder, the dots or spaces plainly longest in direction 

 parallel with the median line, proving the longitudinal to be finer 

 than the transverse lines. 



One obstacle in the way of resolution of the longitudinal striae 

 is the presence of diffraction lines. The valves being so narrow 

 increases this difficulty. Only after much time is wasted, and 

 after many discouraging failures, will the patient observer receive 

 the reward of success. 



Nitzschia curvula Sm. — The unusual number of spurious 

 appearances in this object leads me to suspect that it possesses a 



* 'American Naturalist,' July, 1873, p. 443. 

 t 'Lens,' vol. i., p. 197. 



X See I. E. Smith, in tlic 'Lens,' April, 1873, p. 115. See also the 'American 

 Naturalist,' May, 1873, p. 316. 



