Riveted ships were "caulked," although not quite the way of old wooden vessels. Sometimes bituminous materials were placed in the faying surfaces between the plates. This worked well until those materials dried and cracked. A later technique was to splay out the exposed edges of the top plate against the underlying plate. This was done with a cold chisel and hammer or latter an air tool. A watertight fit could be obtained in this manner.
There are many ways of plating. Seams run horizontal. A line of horizontal plating is known as a "strake." Strakes were usually overlapped on an in-and-out basis. That is alternating strakes would appear to be "on top" from the outside. There were other plating schemes, but in-an-out was the easiest for the yard.
Butts are the vertical joints between plates. These can be made in a variety of ways, just like seams. One common way is to put a "butt plate" on the inside. This is a smaller piece that covers the butt and to which both adjoining outer plates are joined. Above the waterline the butt plates were sometimes placed on the outside of the hull to provide a smoother interior.
Shell plates (the big exterior plates) could also be formed with a sort of recess to accept an overlapping plate. This was more expensive and time consuming.
A major problem in plating a ship was to make sure the butts were staggered. That is, one butt joint could not be in a vertical line with the joints on the plates above and below. In fact, you wanted several strakes between aligned butt joints. So, great care was used in calculating the size of plates to avoid "overlapping butts." (I can hear the chuckles now.)
The big problem with plating a ship was getting the plates to lie flat against the underlying framework. As near to 100% contact between vertical frame and plate as possible was desired. In-and-out plating caused gaps the thickness of the plate at the "out" strakes. American yards often inserted spacer pieces as needed. British yards more often "joggled" the frames. That is, they made the frames shaped to fit the in-and-out plating. This was more expensive and time-consuming, but joggled ships were stronger.
Techniques used to plate ships varied not only from yard to yard, but also from country to country. As David H. points out, it was often possible to tell the country of origin of a ship by the way its plates were put together.
The working of a ship in a seaway as David H. so vividly described caused rivets to loosen and begin "weeping." (water came through) In Titanic's day (before welding) these rivets were removed and new ones fitted. Later, it was common to run a bead of weld around the head to seal this sort of minor leak.
Steel shipbuilding in Titanic's time was really an extension of wooden ship construction. That is, frames were set up just as in days of yore and strakes put hung to form the hull. The result was a very tightly woven basket. Like a basket, the individual pieces (reeds in the basket, plates and frames in the ship) remained discreet. Only during WW-II did welding begin to change things. Today's welded ships are best thought of as one-piece hulls.
What has been overlooked even by modern scientific analysis of Titanic's accident is that the rivets in the seams were not the only fasteners which held the plates together. Each plate was also riveted to the underlying frames. Thus, if a seam rivet was knocked loose, the plate remained in poisition and the seam might not open at all. I can show proof of this on an existing 1911 hull. The strength and watertightness of a riveted hull comes from more than just the butts and seams. It is the totality of the construction--just like a well-made basket.
One-piece containers break under strain while baskets can bend and "give." Well into living memory many sailors avoided welded ships in favor of riveted hulls for this reason. The idea was that the riveted hull would be more able to hold together because of the nature of its construction. And, there was some truth to this. Many early welded ships broke apart at sea. This was the case of many Liberty ships during WW-II.
Back to the butt joints for a final note. On Titanic they were arranged so that any vertical edges faced aft. This was thought to provide a "smoother" surface. Only later with the development of aerodynamics was it discovered that this method actually created "burbles" in the water flowing past the hull. These burbles rrobbed speed and increased fuel consumption. It was learned that butt seams should be reversed. Unfortunately, by then welding had become the way of shipbuilding.
--David G. Brown