ZOOLOGY AND BOTANY, MICROSCOPY, ETC. 767 



•wonder that, while it is the most perfect and the most fruitful instru- 

 ment of precise research, its own announced results oft need verifying. 

 It is not strange that theories have been put forth, discoveries pro- 

 claimed, by observers young and old, which later and more careful 

 researches have failed to substantiate or have entirely overthrown. 

 How many theories have been advanced in regard to the nature of the 

 diatoms, the structure of their frustules and of their living substances ; 

 as to their mode of motion, and even as to their place in the broad 

 classification of biology. Are they animal or vegetable, or alternately 

 one and the other ? There is authority and argument for either view. 

 And when we ask as to the real nature of the regular and beautiful 

 systems of markings their valves display, we can find theories in 

 plenty, but all as yet unverified. They have been declared to be 

 ridges, or furrows, rows of knobs, of areolations, or of minute apertures 

 through the glassy wall of the valve. We have been told we see 

 them most perfectly when they show as hexagons or as circles, or as a 

 wicker basket pattern ; their size and distance apart have been given 

 a most diverse measure. But now we have learned from Abbe's 

 researches how little reliance is to be placed on any of these appear- 

 ances, and how easily many of these various images may be made to 

 appear from a given structural detail by proper manipulation of the 

 light and of the focus. . . . 



But more particularly does microscopic vision itself need verifica- 

 tion. The things seen must not too readily be taken to be the 

 invisible realities. The eye in ordinary vision needs, more than the 

 other senses, to be trained to see aright, and when so trained surpasses 

 all the rest in the fullness of its revelations. So still more does the 

 microscopic vision require careful training that we may be able to 

 safely judge of the reality from the appearances that it affords us. 



Permit me then to suggest in brief outline some of the means 

 which we should employ for microscopic verification, and I would 

 name the most obvious : — 



1. The Eepetition of Observations. — Under varying conditions and 

 by various observations, that which really exists ought still to be 

 seen. It is, of course, not true that every eye can see what the trained 

 adept at the Microscope can easily discover, or that a rare form can be 

 seen again whenever desired. But in general, what man has seen that 

 man can see again, and unsupported discoveries must always be 

 regarded as doubtful till they are verified by repetition. One who 

 has any experience in microscopy can at least see what is in focus 

 beneath the instrument, when his attention has been once called to it, 

 ■vyhile on the other hand, even the well-trained eye is in danger of 

 projecting the mental preconceptions of the observer into the focal 

 plane of the objective, and seeing in the object under examination, not 

 what is really there, but what some theory demands shall be. Spectral 

 lines seem real to one observer that are easily rejected by another, 

 and images seen under one set of conditions disappear under another, 

 and their presence or absence can be accounted for. One of the 

 greatest benefits of a Society like this, and of the smaller local associa* 

 tions, is that they afford opportunity for this comparison of observa- 



3 E 2 



