By start off large, you are likely thinking of Meganeura (see by clicking on the following link click-hereof the Pennsylvanian as an example. There were also very large millipede like critters (see arthropleura by clicking on the link).

The large size of these organisms is related to the fact that atmospheric oxygen (currently at 21%) ranged from 35 to 38% of the atmosphere at this time. Oxygen is heavier than Nitrogen and thus made a denser atmosphere, allowing larger insects to fly at that time, hence Meganura and other "big" insects. Of course Arthropleura does not fly, but here the greater availability of oxygen may have allowed larger body size.

However another feature needs to be considered with Arthopleura. An exoskeleton is heavy relative to an endoskeleton capable of supportng an organism of the same body weight. Thus Arthropleura would be very slow moving. This is great if you live in an environment without predators. The rise of Oxygen also corresponds to the time that vertebrates commenced to invade land, and by the late Pennsylvanian land was a dangerous place with several fairly well-adapted terrestrial vertebrate carnivores. While it has not been explored in detail, one can only surmise that Arthropleura became a prey item (I disagree with the Wikipedia entry --- click here ---- that states " On land an adult Arthropleura would have had few enemies." That would not be true by the end of the Pennsylvanian. I also disagree that it could move "quickly" I suspect it was slow and ponderous due to the exoskeleton. )

You can imagine evolution as a very slow game of king of the hill. On a surface with many peaks and valleys, the species with the best chance to survive are the ones closest to the top. Over time, reproduction and mutation allow species to move along the landscape to climb higher and higher. In this representation, an organism at the top of the surface is the best it can possibly be at surviving in its particular environment.

However, as you may know, environments change over long time scales. In our model of evolution, you can think of this as the surface slowly moving areas that were once hills sink, those which were valleys might rise. The organisms that are closer to the new hill are now more likely to survive and reproduce, and therefore the race begins again for the new hill. Species have to change in order to survive in their new environment.

Now that we have our model of evolution, lets look at insects. In ancient times, the world was a much hotter place. Large swamps and warmer temperatures didnt require things like fur for insulation that we see animals today. However, as the climate changed, large insects are no longer very good at surviving in our current environment the hill has shifted to favor smaller insects. Evolution and climate change are dynamic processes, meaning they keep operating all the time. Animals, climate, and even people are continuing to change!

I think the best answer is "they didn't." The first insects were not terribly large, and modern day insects aren't really any smaller. A better question is what made it possible for the (now extinct) giant versions of insects to evolve in the first place, why they could survive at the time that they did when insects that large would be impossible now? I can confidently say that we don't know the answer, but here are some possibilities:

-Oxygen levels: one of the things that limit insect body size today is the content of oxygen in the atmosphere. Oxygen may have been more plentiful in the past, and the insects may have been larger to compensate for it.

-Competition with vertebrates: the external skeleton that insects have is ideal for making small animals, but for structural reasons an internal skeleton like that of vertebrates is much better for larger animals. When insects first exploded onto the scene about four-hundred million years ago, there were no vertebrates on land, and arguably none even in the ocean (there were things called jawless fishes, which were the ancestors to vertebrates, but they did not have the vertebral column that characterizes the group). When vertebrates began moving onto land, they may have supplanted the giant insects - an insect weighing a few milligrams may be better than anything a vertebrate could do at that size, but an insect weighing a hundred grams might be inferior to a small pterosaur, bird, or rodent doing the same thing.

-Chance extinction: similar to the mass extinction that wiped out the non-bird dinosaurs, there was a great mass extinction that took place about 250 million years ago (in fact, even bigger than the one that did in the dinos), before the first mammals or flying vertebrates appeared. The giant insects may simply have been wiped out, and the animals that arose to replace them aren't any better than they were, but happened to survive the catastrophe.

That's a great question, and one that didn't have an answer that scientists really agreed on until just a couple years ago. In 2007, some scientists found that insects were really big about 300 million years ago probably because there was more oxygen in the atmosphere. It was suspected that higher amounts of oxygen caused these giant insects, but nobody was really sure until this research came out. Insects don't have lungs like us, but use a system of air-filled tubes to breathe. These tubes guide the oxygen directly to all their cells. But insects can only have so much space for these tubes in their bodies because they need room for all their other internal parts. But 300 million years ago, the Earth had about one third more the oxygen it does now, so the air-filled tubes could be filled with a higher level of oxygen and didn't have to take up more space, and this allowed the insects to get much bigger than they are today. When the levels of oxygen decreased, it's thought that these giant insects couldn't survive and died out or evolved into smaller insects. To read more about this, check out these websites:

Not all of the insects of the past were big, and they probably started off small, but there were some huge ones about 200 to 300 million years ago. One species had a wingspan of over 2 feet (71 cm). There were also many small insects. So the real question is, why don't we see the huge ones anymore? It's not because insects were not successful. They live in every habitat today. Their numbers are huge. There are at least a million species of insects, filling all sorts of roles in the ecosystem. The answer may be in the way they breathe. Like all animals, insects need oxygen. Each cell needs a constant supply. Insects that live on land usually breathe through a series of dead-end tubes along their bodies. This is called the tracheal system. We pump air in and out of our lungs with muscles. In insects, the air just sort of oozes through the tubes on its own. Oxygen moves from areas where it is high (the atmosphere) to where it is low (the tube and the tissues and fluids around the tube). This process of things moving from areas of high concentration to low concentration is called diffusion.

You can see many examples of diffusion. Take a clear container of water and carefully put in one drop of food color, trying not keep things as still as possible. You will see the color spread out very slowly. What is happening is that the water molecules are moving around randomly. As they do, they bump into food coloring molecules. This random bumping spreads them out. Random movement of the molecules in the air works the same way.

Using diffusion to supply oxygen is good for the insect because it does not require any energy. It is bad because it is slow. As insect sizes increase, the length of the tubes increases. It takes a lot longer for the air to travel in a longer tube and supply may not keep up with demand. When the huge insects were around the level of oxygen in the atmosphere was roughly 75% higher than it is now. So one hypothesis about why some were so big is that the tubes were efficient enough when there was more oxygen in the air.

Scientists look for evidence to support their hypotheses or to show that the hypotheses cannot be true. Some evidence that supports the oxygen hypothesis is that when oxygen levels are low, they do not grow as big and some evolve to have a smaller body size. Insects grown in atmospheres with less oxygen have larger tracheal systems and those grown in atmospheres with more oxygen have smaller tracheal systems. This suggests that insects only have a tube system as big as they need. There may be a cost to having the tubes, such as water loss, infection, parasite invasion, etc. It may be that the tube system can only get so big. There are other hypotheses too. Birds and bats were not around way back then. Maybe they were better at doing what these huge insects did and drove them to extinction through competition. What sort of evidence would support this hypothesis?