I consider it essential that the photographer should do his own printing and enlarging. The final effect of the finished print depends so much on these operations. ↗
To consult the statistician after an experiment is finished is often merely to ask him to conduct a post mortem examination. He can perhaps say what the experiment died of. ↗