Can you provide the full prompt? The posted images only have what appears to be a partial prompt with text cut off on both top and bottom.

This seems like only half the outfit? Women also wore knickers, as seen here. Can you confirm there were any images like these in the training data?

Having analysed your otherwise excellent LoRA with my SafeTriggers tool;

https://civitai.com/models/1343150/safetriggers-python-tool-for-reading-tags-and-triggers-from-loras

it appears the phrases your LoRA is trained for are as follows;

safetriggers; ./FLUX1/VictorianEdwardianUnderwear-f1-v000003.safetensors trigger words and phrases are;

OrderedDict([('likely from the early 20th century', 11),

('this is a sepia-toned photograph from the early 20th century',

4),

('this sepia-toned photograph', 4),

('white', 3),

('likely cotton or linen', 3),

('lace-trimmed', 3),

('lightweight fabric', 3),

('knee-length', 3),

('likely from the late 19th or early 20th century', 3),

('sleeveless', 2),

('gray', 2),

('this is a sepia-toned', 2),

('this is a sepia-toned photograph', 2),

('slightly shiny fabric', 2),

('featuring intricate', 2),

('this is a sepia-toned photograph of a woman from the early 20th '

'century',

2),

('the image is a sepia-toned', 2),

('cinched at the waist', 2),

('dark', 2),

('is dressed in a vibrant', 2),

with the rest appearing only ONCE.

There appear to be no shorter triggers stored in the LoRA.

I hope this helps.