Just as the part of the result dominates the available space, so is forming the pictures of the results the backbone of the work. Any results that you can not get convincingly documented, should not even be described. Use only those pictures whose quality you consider to be excellent (the others are getting things worse rather than better, when you come to think about it).
If you have been advised by your supervisor well, you have already taken into consideration, from the very beginning, to gain the optimum of each test series images / registration curves, etc.
If not, you can now find it necessary to repeat some tests just to get viable photomicrographs (no dust particles on any sections / with regular and even illumination / with important content placed in focus), derivatives (without askew parts / without power-related artifacts / with good reversibility), or blots (without ‘snaking’ the gel / without colorings / with standards properly covered).
That is so annoying and that is so inevitable; if your best pictures (and these are exactly the ones that belong in the dissertation!) make a negligent impression – guess what the experts will think about the reliability and credibility of your experiments!
Illustrations in a dissertation may and should be of large format (the sharpness is increased by Photomicrographs’ requirements) – http://www.a-mentor.co.uk/services/research-writing/dissertation-proposal/. They are the focal point for the casual reader, and together with the corresponding figure legends, they occasionally must, if necessary, carry the whole “message”.
So, think in particular on the part of your audience that is not familiar with the subject matter of your work! All figures must be served “bite-size”: superscribe everything you will see on the picture, or you will need guidance for. On photomicrographs there must be the essential cell types (or components thereof), or accordingly – layers, of tissue as they are called; something that also is essential for your message is to contain file shares highlighted – arrows, asterisks, etc. (and of course the legend should say so).
Marks of the zero lines and the time periods in which a test substance was applied also should be found on pictures of registration curves, as well as scales or calibration bars for the time and the measured parameters.