No--look at any scientific method. PART of it is experimentatio--but all of hte versions of it have some observation, maybe not visual, but one of the 5 senses.
You are correct. However your logic does not take into account the full range of experimental types. Evoultion could be studied via extremely long term
natural experiments - for example if we managed to catalogue say 99% of the species on the planet we could expect if evolutionary theory holds that new species would evolve over time. If we continued to catalogue species over hundreds of thousands of years we should see new species - more than could encompass the 1% that we missed before - and that these species would be similar to the initial batch in some ways but different in others(and eventually very different if the observations were continued long enough). The point of experimentation in any scientific method is to satisfy the falsifiability requirement - evolution is falsifiable, but to falsify it takes a lot of time.
Evolution theory surely isn't an exacting science like physics, but it is based on empirical observation and data gathering. It's not like someone just came up with the idea based on nothing. The scientific community generally accepts evolution theory as science.