quired. Thus a method such as a colorimetric determination of alkalinity by 

 standards, suffers from our mental inability to superimpose a grid of color 

 standards between the provided standards and to interpolate the actual value. 

 Such a method soon gives way to a visual linear presentation. Hence, audible 

 records are "analyzed" for presentation to our sight, hardnesses are shown on 

 a scale of a hardness tester, and in every way the burden of presenting quanti- 

 tative evidence to the mind is thrown upon the innage sense. 



To summarize these rather rambling ideas on the senses, I feel that we 

 are missing some important opportunities in presentation of information. We 

 must consider the special capacities and limitations of the senses. Our visual 

 sense has extremely poor memory for most time sequences and time integra- 

 tions, but is capable of absorbing and recalling innumerable complex three di- 

 mensional figures. It has incredible capacities to recognize shapes, it can re- 

 call without effort innumerable car body types. I can conceive that an annual 

 record of thermal structure can be absorbed in one effort of the mind, if pre- 

 sented as an isometric view of a three dimensional figure or even as a "thumb 

 movie" of rapidly riffled small diagrams. Yet we present the mind with un- 

 recognizable dots in time and space, or complex tables of data, from which we 

 must mentally derive trends or plot curves. This failure to consider optimum 

 presentation is also conspicuous in primary presentation. When we need to 

 detect correlating pips in a strong random background, we ask the eye to pick 

 out these rare points by memory in time sequence, a task which it performs 

 very poorly. Yet we possess an analyzing mechanism of tremendous power 

 operating between our two eyes that can neglect all non-correlating points and 

 cause the correlating ones virtually to pop from the screen. Can we afford to 

 besiege the mind with tables of data when in one effort a three dimensional ex- 

 pression can be absorbed? 



Lest you are becoming apprehensive that I have forgotten the subject of 

 this symposium, let us consider instruments. 



They are expressions of the fact that man's mind can conceive of more 

 than it can perceive. They are extenders of our senses and link the mind more 

 closely to the universe. But they are mechanisms, and miechanistically we 

 might say that they sample nnatter or energy and present their findings, either 

 quantitatively or qualitatively, to our senses. This presentation may or may 

 not be transiormed in sonme way. Non- transforming sajnplers are called col- 

 lectors, a great class of scientific instruments that are borderline between true 

 tools and instruments. As already pointed out these are of greatest value in 

 the broad qualitative phases of an investigation. They must always retain an 

 important place in an investigation, yet they should evolve in a manner commen- 

 surate with the growth of basic concepts in order for their information to be fed 

 to the mind most efficaciously. Collecting instruments produce much detail and 

 as a consequence tend to stultify an investigation by sheer mass of information 

 The plankton net can be discussed here, for while its information is invaluable 

 for isolated intense, and specific investigations, when the broad descriptive 

 phases of the investigation have been completed, it is essential that methods of 

 analysis be developed that can present the planktonic component of the sea in 

 the least detailed yet most informative manner. This would be a presentation 

 in which the breakdowns in time, dimension, and type were no larger nor small- 

 er than their significant magnitude in the sea at the degree of complexity of the 

 required extension of our concept. For example, if a plankton collection is in- 

 tended to determine available food substances, we should not develop instru- 

 ments and methods for greater than significant spacial and taxonomic resolution 

 but rather integrate the sample spacially over the typified area, and analyze it 

 for total organic carbon, by phyla and size. The resultant information would 



