May 31, 1998 is known for being a widespread tornado outbreak that affected the United States from South Dakota to New England. The outbreak also included a derecho that moved from Minnesota to western New York overnight May 30th into the 31st.
Let's take a look at the atmosphere that produced these storms. (Once again I had issues with data, so I only have maps from 18 UTC on May 31st.)
Strong westerly mid-level flow was present across most of the northern roughly two thirds of the U.S., east of the Rockies, below a deep trough over and west of the Hudson Bay. The Polar Jet phased in with the Subtropical Jet over the northern Great Lakes.
Zooming in to the east coast, and contouring every 2 mb, we see that, although the main forcing for storms was still way back in the Great Lakes, there was another surface pressure trough east of the Appalachians, stretching southward from south-central Pennsylvania, through northern Virginia, western North Carolina, central South Carolina, and southeastern Georgia.
0-3 km Helicity values were above 200 m2/s2 for much of Pennsylvania. An area of 350 m2/s2 cut out part of northeastern Pennsylvania.
Although there were many tornadoes throughout the Northeast and Great Lakes, we'll only look in detail at the ones that affected the Mount Holly CWA.
This case is interesting because it features tornadoes from two different storm modes. Initially, a line of discrete cells stretched southwestward from the Poconos to Lancaster County. Storms in this line produced this first tornado, an F3 in Blooming Grove and Pecks Pond, Pike County:
This Level III data is much coarser than today's. You can clearly see an 81-knot couplet in Pike County at 2313 UTC (7:13 p.m.), associated with the strongest discrete cell. There were a few weaker tornadoes in Luzerne County this day as well; the rotation you see in the storms near Wilkes-Barre is likely related to those.
If we zoom out (and move forward in time about an hour and a half) you can see the broken line of storms moving southeastward towards the Delaware Valley:
Notice the strongest storm, in Berks County, has a 90-knot couplet associated with it on Storm-Relative Velocity. At this time, an F3 tornado was moving through that area.
The couplet on radar is seen here at 0058 UTC (8:58 p.m.) southeast of Shoemakersville, northeast of Leesport, and northwest of Fleetwood. The bright yellow line running southwest-to-northeast is U.S. 222.
This line of storms weakened as it continued moving southeast, finally hitting the city of Philadelphia around 11 p.m.
But a large squall line was just entering from the west at the same time.
An F2 tornado struck near Quarryville in Lancaster County with the squall line.
You can see some broad rotation in the area of southwestern Lancaster County, but nothing obvious. The lack of an obvious couplet may result from the storm's distance from the KDIX radar, however.
An F1 tornado touched down in Leanpe, Chester County, south of West Chester, but was also not obviously visible on radar.
The last tornado of the night was an F2 that moved through Northeast Philadelphia and southern Bensalem at 0521 UTC (1:21 a.m.):
You can see the Roosevelt Boulevard and Woodhaven Road on the map. The tornado moved along Southampton Road, heavily damaging industrial areas.
The National Weather Service has a few pictures of the damage in Lyons, Berks County as well.