Weather data from the early twentieth century are hard to come by. There were no weather satellites, no radars, no upper-air weather observations, and very few surface weather observations. But from those few surface observations, and other news accounts of the time, a picture emerges which suggests that May 10th was only one in a series of active, stormy days in the Plains in early May 1905.
Another devastating tornado struck the central Plains only two days earlier. The town of Marquette, Kansas was struck just before midnight on Monday, May 8, resulting in 34 deaths. News articles include reports of other tornadoes striking during the afternoon and evening of Tuesday, May 9, just southwest of Gotebo (30 miles north of Snyder), near Ringwood (west of Enid), and even in St. Joseph, Missouri. One article stated, “Cyclones seem to be general this year, evenly divided throughout all the states.” 
Synoptic weather maps show the general surface weather patterns from North and Central America westward across the Pacific Ocean on the mornings of May 9th, May 10th, and May 11th.  (Times are given as 1300 Greenwich Mean Time, or GMT; subtract six hours for Central Standard Time, so the charts are valid at 7 AM CST.) These charts, which were analyzed by hand, show the positions of fronts and other surface weather features. Daily weather maps over the contiguous United States, also valid at 7 AM CST on May 9th, 10th, and 11th,  (the listed time of “8 A. M.” is most likely Eastern Standard Time, one hour ahead of CST) provide more detail of pressure and temperature, but do not identify frontal positions.
Focusing attention on the area of the United States, we find a large area of low pressure over the Rocky Mountains on the morning of May 9th, with one center over Wyoming and another over southeastern Colorado. A front can be analyzed from Virginia westward into Nebraska, and most likely is lifting north as a warm front in the central Plains. Early-morning temperatures are near or above 70 south of the front, from Kansas southward. South winds prevail throughout the central and southern Plains, east of the surface low, indicating that moist air is moving north across the area from the Gulf of Mexico. This is still more than 36 hours before the Snyder tornado.
On the following morning, May 10th, we find that a low center has moved to eastern South Dakota, while another has become established over western Colorado. It is at this point that our meteorological knowledge must be called upon to infer what is happening in the upper levels of the atmosphere. Since weather systems generally move from west to east, we conclude that an intense low-pressure trough aloft, which created pressure falls over the Rockies the day before, has moved northeast and caused pressures to fall over the northern Plains. Hence, the low over South Dakota most likely tracked from southeastern Colorado. This weather system likely initiated the tornadoes on May 9th, mentioned earlier as having occurred from Oklahoma to Missouri. But the lingering area of low pressure over Colorado on May 10th implies that another upper-level trough is dropping into the Rockies behind the one that is lifting into the northern Plains. (This trough is known in the weather business as a “kicker,” since it likely kicked the lead trough out into the Plains ahead of it.) Although areas of Kansas and western Oklahoma and Texas are cooler on the morning of the 10th, winds remain south in these areas in response to the new developing area of low pressure to the west. It is likely that this “kicker” trough was the storm system responsible for the tornado outbreak on May 10th, which would level Snyder roughly 14 hours later.
To see what happened next, we must jump ahead momentarily to the next data set, on the morning of May 11th.  We find a deep surface low over northern Kansas, a warm front extending eastward and lifting north through the middle Mississippi Valley, and a cold front southward from the low into Oklahoma and Texas. To the south of the warm front, and east of the cold front, south winds are continuing to draw Gulf moisture northward – and most likely have been doing so for most of the past two days.
What happened during the 12 hours or so prior to the Snyder tornado now can be inferred. The low pressure center, which began the day over western Colorado, likely moved to, or reformed over, the central and southern High Plains, around southeastern Colorado or the Oklahoma Panhandle. This is a typical evolution in most Plains-type severe weather outbreaks, and indicates that an intense upper-level disturbance (the “kicker” trough) dove into the central or southern Rockies during the day. South winds increased over Texas and Oklahoma in response to the developing low to the west, drawing warm, moist, unstable air back north into the region. That evening, the disturbance continued eastward and emerged over the Plains, lifting the unstable air and initiating severe thunderstorms.
It is virtually impossible to prepare a smaller-scale meteorological diagnosis of this event based on the available data. So we cannot say why southwest Oklahoma, and Snyder in particular, became the targets of nature’s wrath on that day.