Geographic isolation is generally thought to be an impediment to growth. While I do not give too much credit to cross-country regressions, they have consistently shown that being landlocked or not have direct access to navigable waters is bad for national income. The reason is that being cut off from trade routes, you do not benefit from technological advances as early as others.
Quamrul Ashraf, Oded Galor and Omer Ozak argue that this logic does not hold for development in prehistoric times. They compute an isolation index by measuring the average time it takes to travel from a capital to any other location in the known world (excluding the Americas and Sub-Saharan Africa). It ranges from 5.5 weeks in Georgia to 12.1 weeks in Malaysia. In the premise that Malthusian economies strive to increase population, not per capita income, they then look at the impact of this isolation index on population density in the years 1, 1000 and 1500. They find that the effect is there, and in fact it is still present when using these antique isolation measures on modern day income per capita.
The big question is now to understand why in old periods isolation was beneficial, for example for China. Is it because isolation made one less susceptible to war and envy? Is it because isolation prevents slavery? Or is it because of a fluke in the data? Note that population density is used as a measure of development, which make for example that Egypt scores particularly low. But population is extremely dense around the Nile... In any case, this paper raises more questions than it answers, which is good.